Edward Thompson: Wartime C.M.E. Discussion

What period was Durrant writing about? That's not how I remember it in the late 50s and early 60s as I shuffled about between KX and Wood Green on a couple of platform tickets or no ticket at all.

This is close to what I remember. 60007 at Alexandra Palace (Wood Green) in 2008:



Syncopated maybe but not 2½ cylinders!

 

Re the 2 : 1 valve gear, bearing in mind that there would have to be a certain amount of dismantling for outside cylinder piston valve examination, was it really so difficult to do a little more by way of pin and bush replacement while it was apart? I just can't help feeling that the 'maintenance difficulties' have been somewhat overstated. I might point out that when on BR in the mid-1960s I worked for a time with Arthur Taylor Jnr. (son of Driver Arthur Taylor of 1935 A4 fame) at one point Foreman Fitter at Kings Cross, and don't recall any adverse comment from him on the 2 : 1 gear.

Shifting sideways for a moment, I feel that it was a shame that (particularly) A4s were never fitted with roller-bearing axle boxes, at least for the drive axle. This would have made a good loco an even better one.

 

Thanks for the video clip. A hefty dose of nostalgia there.

 

Can someone confirm the title of this book please? I have a copy of most books that cover the LNER Pacifics but do not recognise Durrant as an author that has covered them....

Thanks....

 

Swindon Apprentice by A. E. (Dusty) Durrant ISBN 1-870754-10-7 by Runpast Publishing 1989.

 

Its Durrants autobigraphy of his apprenticeship days, which includes much material about his trips all over europe loco watching, and his sketches of the (often gigantic) locomotives he felt that BR should have been building. The paragraph quoted goes on to say "At the time I considered them [the Peppercorns] the best express locomotives in Britain, an opinion largely unchanged today."

I don't suppose that Durrant meant that the Gresley locomotives were running on exactly 2.500 cylinders. However this is what the Cox/Stanier report has to say (extracted from the earlier post)

There's absolutely no point in my listening to a single video clip and saying whether Durrant is right or wrong even if its representative. I lack the trained ear. However *if* my understanding of the conjugated gear is correct (VERY big if) the exhaust ought not to be syncopated if the valve events are as even as they theoretically ought to be, so it seems to me the syncopation you observe must be a sign of valve events that are sub optimal. But I know very little of this and am happy to be corrected by those who understand the engineering challenges.

Its quite clear from his book that Durrant had a very open mind: he praises aspects of French and German practice, and is often very critical of things in Britain, including those of the line he trained on.

 

The conjugated valve gear was controversial long before Thompson came onto the scene.

What are probably OVS Bulleids recollections are captured in his son's Book "Master Builder's of Steam" of the mauling Gresley received when he delivered his paper "The Three Cylinder High Pressure Locomotive" to the I.Mech.E. at Newcastle upon Tyne in July 1925. The following quote is from the book:

"He fell into the unexpected error of over-playing the case for the three-cylinder arrangement. After pointing out how his conjugated lever saved a set of valve-gear and thus removed the main objection to three cylinders, and supporting this by examples of similar gear used in Germany and the USA, he went on to summarise the advantages:-
1. Less coal consumption
2. Increased mileage between general repairs
3. Less tyre wear
4. Lighter reciprocating parts, consequently reduced hammer-blow.
5. More uniformed starting-effort
6. Lower permissible factor of adhesion
7. Earlier cut-off in full gear.
Of these item 7 merely repeats 1, 6 repeats item 5, item 3 is a detail of 2. Moreover item 4 sins by obmission: with four-cylinders hammer blow can be nil. Frankly the case for three cylinders was marginal, and Ivatt was probably right once and for all when he advised using the least number of cylinders that would do the duty. But when a leading designer overstates a marginal case it shows emotion at work - not necessarily a bad thing so long a speaker and audience recognize it. Gresley fresh from seeing a Castle walk away from a pacific could not resist a vague gesture of defiance in his penultimate paragraph: "Undoubtedly a four cylinder engine can be designed, the power which will exceed that of a three-cylinder with the same loading gauge limits, but the construction of such an engine at the present moment would be premature, in the same way as the construction of three-cylinder locomotives nearly eighty years ago was unnecessary for the requirements of the times." There was a lively discussion. Clayton (Maunsell's Assistant) said that the two-to-one lever had the inherent defect of lost motion, hence inadequate control: Gresley replied that the trouble had been overcome by the use of grease-packed ball bearing joints. Beames challenged the statement that "With the present type of locomotive boiler it is neither practical nor economic to make any considerable increase in boiler pressure." Sir Vincent Raven, President, who was in the chair, said "I always adhered to the Stephenson Valve Gear because I believe in simplicity, I used three sets of valve gear, and if I went back to railway work today, I should do the same again... I really think the distinct advantage of the three cylinder engine for locomotive purposes has been proved". Holcroft's contribution to the two-to-one lever was also raised, but Gresley, out of character, dismissed it in a cavalier fashion"

I must admit I have some sympathy for Gresley, when Raven - the President of the I.Mech.E. (the professional body) , the chairman of the meeting he was asked to address and one of his predecessors dismisses his work in public - that must of hurt.

I.Mech.E. meetings of that period are not usually this lively and this episode does stand out in a book whose mantra is about how the CME helped each other. Whilst at University I looked up the original paper within the I.Mech.E. proceedings of 1925 in the Library but I was disappointed to find the lecture ended boringly with a statement by Gresley that he does not wish to answer any questions! I suspect someone tactfully switch it for the notes of a follow-up lecture! I am sure the I.Mech.E. Library at Birdcage Walk London will help any researcher with a copy (I suggest you also ask for Stanier's Presidential address "The Position of the Locomotive in Mechanical Engineering" which is beautify written).

Regarding Thompson, I suspect he has suffered of having a bad name given to him from the recollections of subordinates who work for him and collected by Rogers etc (the only witnesses still available). His managerial style did not go down well in 1930s Doncaster or in recollection in the more left wing society views of the post war period, but are not uncommon today if you work for a North American company where macho managerial philosophy and the micro-management/surveillance of employees are fashionable in some industries. He had some idiosyncrasies that RC Bond observed but they can be explained by modern psychology. He might well of been isolated by colleagues due to his unusual background as a direct entry in management as an Oxbridge Graduate rather than working his way up from a premium apprentice (internship) route either through his own making (arrogance) or social awkwardness. I agree with SAC Martin - we need to examine his actions in context: -

1) There was a war to be won and economy of cost and labour was key.
2) Conjugated valve gear was controversial in the industry but discussion appears to have been taboo.
2) Gresley was in post for 18 years as CME of the LNER - not very healthy with respect to succession planning as his direct reports were either similar age or too inexperienced, but this can be linked to the unexpected loss of his chosen successor (Bulleid) and the effects off WW1 where a whole generation of young men were lost (same pattern can be found when you look at the age and time in office of UK politicians in the 1920s and 1930s). After the loss of Bulleid Gresley did not appoint a successor but shared his work with the very young Spencer and Newsome. "His own department tended to be a long way above his staff, and only Bulleid had got close enough for free discussion and the occasional, sobering, terse rejoinder" (HAV Bulleid)
3) Having made some internal enemies in your appointment where you were not the obvious candidate, what do you do? who do you go for advice (Stanier & Cox) and if you ask for advice should you follow it?
4) Gresley was not always easy to work with (again HAV Bullied):
"After such minor battles Gresley would stand comfortably with his back to the impressive fire in his office, and discuss the form with Bulleid. Hunamly, they sometimes discussed their wisdom, and sometimes the stupidity of others. They did not always agree: once Gresley took up his fireside position quite indignantly, wanting to know how it was that none of his assistants ever seamed to come up with suggestions. "By the way," said Bulleid casually, "you remember that draughtsman you agreed to see last week" - "of course I do, the dam' fool" "Well, there you are," said Bulleid "Do you think he'll ever come back with another suggestion?" Gresley was slightly penitent; each found it easier to see brusqueness in the other than in himself.


If you can find it in a second hand book shop or online I suggest HAV Bulleid's "Master Builder's of Steam" is a good book to get hold of: -
Chapter 1 - HA Ivatt (contributions from HG Ivatt & OVS Bulleid)
Chapter 2 - Gresley (contributions from OVS Bulleid)
Chapter 3 - Bulleid (contributions from OVS Bulleid)
Chapter 4 - Churchward (contributions from WA Stanier)
Chapter 5 - Stanier (contributions from WA Stanier)
Chapter 6 - HG Ivatt (contributions from HG Ivatt)

Also a great preface.
"The author of this book , given an advantageous view-point (and using it, as a practical engineer should), has told us that these data an anecdotes and reminiscences have at least some value in illuminating the humanistic art of engineering.
We agree.
In the course of our careers, we have all mulled things over with each other and learned therefrom. Problems shared are problems solved etc."
OVS Bulleid
HG Ivatt
WA Stanier

 

We can exaggerate that: I got pulled up on it recently talking to a WW2 veteran about sailboat history when I blithely said much the same thing and he corrected me (it being his parents generation). It seems the overall fatality rate in WW1 was around 10% of those who served. Grievous enough and disastrous enough by any measure, but a substantial majority of the generation did survive.

BTW Master Builders of Steam is available electronically. I bought it fairly recently but have somehow forgotten where and how!!

 

Whilst what you say about WW1 may be true, there were still many more who were injured, either physically or mentally...

 

Do you mind me commenting? The websites on casualties state that from the British Isles 5.7 million men served in the Army.
in WW1. There were 705.000 fatal casualties of which 41.000 officers. You can calculate the percentages yourself.
It might be that possible future CME's were more drawn to the Flying Corps.
Kind regards
Jos Koopmans

 

Its available for free download here https://archive.org/details/MasterBuildersOfSteam Its not pirated but public domain.

 

Ok, Wall of Text time....

Having looked through the report, I’m somewhat disappointed that the gear failures have not been put in any sort of context.

The remit requests a considered opinion on the merits and demerits of this gear, and surely for a reasoned assessment of a mechanical system that is experiencing problems, the situation at the onset of the problems should be compared with that prior to the appearance of the problems? Cox states in his Loco Panorama that “This (the gear) was a Gresley feature which he (ET) particularly disliked and which under wartime conditions of maintenance was giving a lot of trouble. I could not, of course, report on what it had been like in the heyday of the A4’s(sic) prewar”

I would thus have to ask why he was not able to report on the failure and wear rates during the previous 25 years of, at least satisfactory, and at best brilliant, use of the conjugated gear, to provide comparison with the problems found recently.

I have no argument with the lost motion calculated for the new gear, but there is no explanation as to whether the 3/8” lost motion in worn gear was found during or before the war. Since Cox states that he can’t report on the prewar situation we must assume that this wear was found on badly maintained locos in wartime.

As the increases and decreases in power at the centre cylinder are calculated from this wear figure, it thus seems reasonable to suggest that under normal prewar maintenance regimes such wear and such significant power imbalance would not occur, at least at the late cutoffs used at low speeds which load the bearings the most.

To simply declare a design as fundamentally unsound because of the inherent multiplication of play is an overstatement, bearing in mind the successful use of the gear prewar. That the design cannot provide as precise a distribution of steam as that provided by three sets of gear is not at issue, nor is the effect of valve stem expansion, but the effective use of the gear prewar indicates that an unequal distribution of steam is not a barrier to fast and efficient running. Nor does it seem to be a cause for damage to the big ends, as the report states that the poor steam distribution is unlikely to contribute to bearing damage in a major way.

The notes on bearing size are interesting, though unlike the comments on the big ends, the report makes no recommendations as to whether increasing the bearing size would improve matters.

I thus have to take issue with the comment “incapable of being made into a “sound mechanical job”. Prior to the depredations of wartime, the gear was a sound mechanical job. It was certainly not an ultimately precise job, but it provided acceptable steam distribution for a fleet of fast and efficient locos, together with a several other classes of mixed traffic and freight locos, for a good 20 years.

I also have to take issue with the comment that rapid pin wear is inevitable. Since Cox admits he made no effort to assess the prewar performance he is in no position to make such a statement. Saying rapid pin wear was inevitable under the present (1942) levels of maintenance would have been logical, but as noted above, no context was provided for the failures.

The discontinuation of conjugated valve gears elsewhere is noted, but no evidence is provided as to what type of gear was abandoned. The Holcroft gear had been abandoned on the SR, but it was far more complex than the Gresley gear.

The comment that the locos will need frequent overhaul in the shops appears to follow from the rapid rates of wear. As pointed out above, the rate of pin wear under peacetime conditions seems not to have been investigated, so to conclude that it is certain that overhauls will be frequent appears unwarranted. Again, qualifying the statement that under the extant conditions, frequent overhauls would be required would have been logical, but no such differentiation was made.

Regarding merits of the gear, none have been noted at all, however the elimination of any valve motion between the frames requiring daily maintenance has surely to be seen as a merit – as has the ease of access to the conjugated gear itself. Replacing the bushes in the gear is unlikely to require a full overhaul, but no mention is made in the text of the amount of work needed to replace the bearings.

All in all I think the part of the report dealing with the gear is at best misleading and at worst heavily biased. Strangely enough, a much more balanced approach, pointing out that the problems were due to wartime effects, noting how uncertain the future was in terms of a return to peacetime maintenance, and suggesting that whilst the end of restrictions was not in sight abandonment of the conjugated gear in new build would be wise, would have been a far better way to proceed.

Asking whether ET was justified to change the type of gear in the light of the report is moot. We know ET wanted to change the gear, Cox says so, so I doubt anything short of a glowing recommendation of the gear would have dissuaded him. As to whether the directors should have allowed ET to change the gear on the basis of the report is surely far more relevant. Having agreed to back ET if a condemnation of the gear was provided by Stanier, I don’t think they had a choice. I suspect on reading the report they looked around at the photos of Mallard, remembered the high speed services running before the war and muttered the 1942 equivalent of wtf? but unless they wanted to engage yet another expert, and possibly have to find yet another CME, I don’t think they would have had the stomach for it in the middle of the war.

Finally, there’s the question of why the report is only now being circulated in its entirety?

Why was Nock was foisted off with such a vague synopsis by (presumably )ET ?

Firstly, as I have suggested earlier in the thread, Stanier and Cox would have have been severely criticised for such a comprehensive trashing of the valve gear used to such effect on the celebrated Gresley Pacifics, including the world speed record holder. The complete lack of any sort of acknowledgement that the problems were recent and war related would have seen to that. Considering the recent speed record tussle between the LMS and LNER, Stanier would have no doubt also been accused of sour grapes.

Secondly it seems likely the report was requested on the basis that whatever was written would go no further than the Directors.

Why was Cox so vague about his input in his Locomotive Panorama?

Writing in about 1964, he had seen the relatively mediocre performance of the Thompson Pacifics and had seen the slow but steady improvements to the performance of the conjugated gear fitted A3/A4 classes, culminating in the grand finale on the ECML. I suspect he was somewhat embarrassed by his input to the report (if I were him, I certainly would have been.)

I’m not sure just how much of the conclusions were included in the Stanier biography, but it appears that the comment about “sound mechanical job” was removed, possibly for the same reason Cox was so vague.

As to whether the report was tailored to suit ET’s needs (and he really did need a categorical condemnation of the Gresley Gear) - I will leave it to others to decide on their own.

 

Lplus, and houghtonga,

many thanks for those (IMHO) very fair objective assessments and further information.

what ever way you look at the report and what happened afterwards, it doesnt provide the justification for (or exonerate) what Thompson did subsequently - and in the case of the middle big end bearing failed to do.

cheers,
julian

 

IIRC the syncopation is due largely to the cranks not being exactly 120º apart to allow for the inclination of the middle cylinder. I need to dig out a book to confirm this but it's been a very long day up at the NVR so it will have until tomorrow.

 

Picked up a copy today while on duty at Wansford. Looking forward to getting to grips with it.

 

Thanks.

Regarding the bearing, It isn't clear whether the changes planned for the big end design which the report appears to approve included the final design, though this quote under "testing and performance" from here http://www.steamindex.com/locotype/grespac.htm by W D Harvey suggests not, as the final design was introduced in 1947;

"Regarding the middle big-end about which so much has been written in recent years, the evidence produced at the depots we visited confirmed my opinion that the semicircular strap was weak and was flexing under load, thereby distorting the brasses and causing these to nip the journal and so cause heating. Proof that flexing occurred was afforded by the fact that when a centre big-end was taken down for examination the jaws of the two brasses were invariably bright and polished where one had fretted against the other – likewise the surfaces of the bronze shims or gluts used for adjustment. More convincing proof was the fact that whenever a brass was found broken in two, as happened occasionally, it was always the back brass that broke, never the front brass, although the front was weakened by a %in keyway'" preventing the circular brasses from rotating. The polished) appearance of the fractured surfaces caused by fretting was convincing proof that the strap flexed. The design of the middle connecting rod and strap closely resembled on a larger scale that in a high class automobile engine and was a magnificent piece of forging and machining. Possibly its designer had such a prototype in mind but overlooked the fact that whereas an internal combustion engine was single-acting and the strap therefore carried little or no load, in a double-acting steam engine this was not so, for the loading was the same on both brasses, for which the strap as designed was ill-suited, being at its weakest at the point where it needed to be the strongest.
I had been much impressed by the excellent design of rod ends that I had seen on German locomotives. These were of T-section with a deep crescent-shaped rib or web at the back in order to resist deflection. This was the design that we recommended and which was ultimately adopted as the standard - not quite as we wanted, as the rib was made concentric instead of crescentshaped, but still a good deal stiffer than the original design.
This improved pattern may be seen on nearly all the Gresley three-cylinder locomotives that have been preserved. Prior to this, several alternative designs of big-ends for three-cylinder locomotives had been produced under Edward Thompson's direction but those seemed to be singularly unattractive in design, - being heavy and clumsy. An interesting aspect of our enquiries was the diversity of opinions expressed on minor points of design, troubles experienced at one depot with certain fittings being almost unknown at another. "

To be fair to Thompson, the report seemed happy with his proposed changes. It is true however that the changes to the bearing shells made by Cook were first suggested by Stanier in the report.

Interestingly the same passage quoted above includes the statement that the worn valve gear was to blame for the bearings overheating. Unfortunately it isn't clear when this investigation was made.

I believe the figure is 127 degrees. The centre cylinder is at 8 degrees and the outside cylinders are at 1 degree, so the centre crank is at 127 degrees from one outer crank and 113 degrees from the other. In fact, this should prevent syncopation from occurring, and on properly set and maintained valves it shouldn't be really noticable. The inner valve lap is, I believe, slightly less than that for the outer valves, possibly to take account of the slight play and some whip at speed, so this in itself may add a slight variation to the valve events.

Edited (many times) re crank info!

 

Interesting. I've certainly read that this uneven crank spacing was cause of the syncopation. I will say that under Bill Harvey, 4771 always sounded quite even.

 

There is a discussion in Holcroft in connection with the N1: as I recall, the variation from 120deg is equivalent to the inclination of the middle cylinder (if the outside ones are level) so as to achieve even valve events.

That layout was worked out theoretically by Holcroft specifically to result in even beats, so I suspect that the famed Gresley syncopation has some other cause. The variation in events caused by a combination of wear and expansion of the valve spindles would presumably change the timing of the middle cylinder relative to the outside cylinders.

Incidentally, the relevant pages in Holcroft are worth reading to get one account of the genesis of the valve gear, and also the extent of the discussions about it between Holcroft and Gresley.

Tom

 

Something that has always puzzled me - why was 1471 (the other original GN A1) not considered for preservation? If 1470 was not available it would seem the obvious choice - remembering that Flying Scotsman was also not considered and would not be here if it was not for Alan Pegler?

On the valve gear issue, the problem is that the cure, the third set of valve gear, was not effective - it was neither easier to maintain nor did it prevent hot big ends. (According to Peter Townend, anyway). So there is a criticism of all three LNER CME's in that the big end design, apparently the real problem, was not satisfactorily addressed. And on that basis the Cox report, which clearly mentions the big end issue, was either ignored or misunderstood by Thompson once he got his permission to build 'his' kind of Pacific.

However, I also quite understand the irritation of someone who has been bypassed for command, on attaining it, it be told by the board, as Thompson apparently was, that there is no need to design anything of your own, just carry on building your predecessor's designs. Especially when you are well aware that said designs are not as perfect as the board seems to think.

It's a complex issue - with a war on there is clearly an issue with developing vanity projects like new designs of Pacifics just to show you can do it differently (and , unfortunately, somewhat less effectively, at least as far as maintenance costs were concerned) .

 

Some two cylinder engines don't either, notably the LNWR 0-8-0's. It's not as critical as people sometimes seem to think. Some 2-cylinder locos operated quite happily with worn cylinders bored to different sizes, for example. And an out of tune Gresley tended to do more work on the middle cylinder than the outside pair, not less as the Dusty Durrant quote implies - this was supposed to be the cause of the big end failures until it became apparent that they continued even with the more even distribution of work obtained from the three sets of valve gear on the Thompson and Peppercorn Pacifics.

There is no engineering reason - though there may be aesthetic ones - why a multi cylinder engine should not have inside cylinders of different dimensions to the outside ones, and then they certainly would not have an even beat.

Engineering is sometimes a voyage of discovery - its an embarrassment to all us LNER enthusiasts that it took a GW man to fix the big end problem - but having said that, when debugging a problem, if you have initially identified the wrong line of enquiry it is terribly difficult to go back to first principles and start over, especially when the data is patchy and capable of multiple interpretations.