Edward Thompson: Wartime C.M.E. Discussion

Good point - why indeed?

 

This thread has become preoccupied with the merits and demerits of conjugated valve gear, but I have a question regarding the P2s.

The P2s were introduced to remove the need for double heading on the Aberdeen line. I have read suggestions that they should have been reallocated south of the border to take advantage of their hauling abilities for the war effort. Maybe one reason this didn't happen was that it would have resulted in a return to double heading, so there was no point?

My question concerns what happened after the rebuilding - presumably double heading did then become necessary again? If so, this increased burden on the fleet suggests to me that rebuilding the P2s was an expensive luxury, that could and should have been delayed until after the end of hostilities. Just a thought.

 

hi smokestack lightning,

i dont know where the rebuilt P2s operated after rebuilding but im pretty sure it wasnt on the aberdeen line so you may be correct! they were built specifically for the aberdeen role.whether this worked out ok, and more pertinent to this thread whether Thompson's rebuild of them was a good idea is quite another matter! simon has his own views of course!
to my mind, the fact that the Tornado team are progressing with an unrebuilt P2 suggests that Thompson did the wrong thing.

incidentally one of the finest 5"g miniature locos built in recent years is a copy of original P2 Earl Marischal by John Heslop, which won not only all the top awards but IMLEC in 1995 (test of efficiency competition). what Thompson did to that design was in my opinion nothing short of criminal butchery with no regard to what might have made the locos work extremely well at minimal cost for a cash strapped LNER. that's the mark of a bad loco engineer in my book (and im no Gresley or LNER fan!)

cheers,
julian

 

A tad unfair methinks. Although the P2s were built for the Aberdeen route they were reported to be prodigious consumers of coal and the Wikipaedia entry for the class notes :

According to B. Spencer the class was rebuilt due to reliability issues during the difficult conditions of World War 2 period, and additionally to take the opportunity to try out a different valve gear arrangement.[28]

Railway author O. S. Nock suggested that Gresley's successor, Edward Thompson, may have made largely unsubstantiated criticisms of the class in order to justify the rebuilding.[26] According to O. Bulleid the class were not an inefficient design, but had been placed into services in which they were underutilised, leading to poor fuel economy.[29]

The LNER webpage however also notes that :


The P2s tended to suffer from a large number of damaged driving crank axles due to the overheating of the big-ends. This was partly due to the stresses at the front end caused by pony truck faults. The same swing-link pony truck mechanism employed on the K3 2-6-0s was used, even though this design had exhibited problems with both the springs and sufficient lubrication. Associated weakening would occasionally lead to failure of a crank axle whilst in traffic. These failures appear to be a major contributing factor to Thompson's decision to rebuild the P2s as A2/2 Pacifics. As a part of the rebuilds, Thompson used coil springs to centre the leading bogies, and it should be noted that none of the A2/2 Pacifics suffered crank axle failures.

After the death of Gresley in 1941, Thompson took over and began his standardisation plans. Due to their small numbers and mixed success, it was almost inevitable that the P2s would be rebuilt by Thompson. Between January 1943 and December 1944, all six were rebuilt as A2/2 Pacifics.

Although often described as a mixed success, the P2s were a bold step forward for British locomotive design. They had a reputation for heavy coal consumption, but this was only really a problem with No. 2001 when it was fitted with its original poppet valve gear. The pony truck had problems, but as with the V2 2-6-2 locomotives, it could have been successfully redesigned. Also, if they were ill-suited for the Aberdeen route as many imply, how come they served their entire lives on the route for which they were designed? It is true that the longer fixed wheel-base did pose problems for the tight corners on this route, but the extra axle gave useful increase in adhesion for this steep route. As with all locomotive designs, the P2s were a trade-off between conflicting requirements.

The final statement "a trade-off between conflicting requirements" is of interest in that those "requirements" would change in response to "operating conditions" and one wonders if Thompson's changes / rebuilds were the result of his interpretation of those changes. Did he interpret them in such a way as to justify his move to independent valve gears? did his use of cylinders between the bogie and driving be based on his father-in-laws (Raven) practice of the same ? did his designs of such extend the frames to a point where they were more liable to fracture ? More importantly - if all locomotive designs involve a degree of compromise what compromises did Thompson have to include in his designs ? was the consequent trade-off better or worse that those of earlier Gresley and later Peppercorn designs ?

Another interesting question; was Thompson's standardisation on the 6'2" driving wheel for his Pacific designs as favoured by Gresley on his V2 design his tribute to Gresley's standards - or an understanding of Gresley's logic for selecting that wheel diameter for the V2 design when the Gresley Pacifics had 6'8" driving wheels hence his use of the last 4 V2s to create his first non-rebuild A2 Pacific.?

And a final point regarding the P2, note that the A1SLT have confirmed that they are looking to (a) review the crank axle design and change it as appropriate (b) have reviewed all components to enable their use on the modern railway and (c) have confirmed that whilst externally it may look like an original Class P2 locomotive its parts will be in accordance with modern day operation.

 

hi fred,
you start by stating i'm a 'tad unfair', then go on to quote a whole host of conflicting stuff which in fact backs up what i stated! what you quoted only casts serious questions over why Thompson did what he did, rather than adopt a cheaper and simpler way of rectifying a few design /maintenance problems. instead he decided on an expensive rebuild programme for a cash strapped LNER!
cheers,
julian

 

I'm not going to comment one way or another on the P2s at present, but I'll throw in a few more lines from Steamindex: http://www.steamindex.com/locotype/gresloco.htm#p2class

"Retrospective and critical
A controversy has arisen about the efficiency of the design. This has partly resulted from Thompson's drastic rebuilding of the class is Pacifics (see A2/2), which implied official condemnation. McKilliop, who drove the locomotives in service and on tests, was critical of the poor riding qualities and high coal consumption. It should be noted that McKillop usually praised Gresley designs, especially the Pacifics. C.J. Allen has disagreed with McKillop, however, (see "Ponies, Pacifics and mikados") basing his argument in terms of performance and power output. O.V.S. Bulleid also had a high regard for the design and was closely associated with the tests performed at the Vitry testing station, near Paris. Geoffrey Lund gave an excellent account of the class in a special issue of Backtrack devoted to the LNER.

Bellwood and Jenkinson (presumably the former) summarized P2 performance: There were insufficient passenger trains loaded to over 500 tons on the Aberdeen route to utilise fully the six locomotives, with the result that many workings were 'over-engined'. The eight-coupled axleboxes did not take kindly to the numerous curves on the route, and with Pacific train loads the P2s were expensive to both operate and maintain.
It is a great pity that their prodigious haulage capacity could not have been used to assist in the working of the enormous wartime loads between Kings Cross and Newcastle. Certainly there would have been ample opportunity for them to be run under near optimum operating, if not maintenance, conditions in relation to power output and adhesion. However, it was not to be. There was no place for Gresley 'super-power~ in the strategy of the new CME and all the P2 class locomotives were rebuilt by Thompson into ungainly Pacifics. As 4-6-2s, they had a very short life on the Aberdeen. road and had all been scrapped before the end of 1961.
Whatever the merits or shortcomings of Cock o'the North, it probably represented Gresley's greatest single step forward and was certainly ahead of its time. It is significant that when, twenty years later, the last express passenger steam locomotive was built for British Railways, it had three cylinders fitted with poppet valves."

"Bulleid, O.V.S. discussion on Spencer, B. The development of L.N.E.R. locomotive design, 1923-1941. J. Instn Loco. Engrs, 1947, 37, 211-12. (Paper No. 465)
Testing 2001 Cock o' the North on the test plant at Vitry. It is interesting that he always refered to the locomotive as "she" as in she compared favourably with the French engines in coal consumption per rail-h.p., and, better still, per d.b.h.p. When tested on the open road between Orleans and Tours it developed a very high horse-power, of the order of 2,800, and was again shown to be an efficient engine from the point of view of coal consumed per d.b.h.p. In service, however, it was an extravagant engine as it was not properly used: instead of working trains well within its capacity over long runs, it was employed on a service such as Edinburgh to Dundee on trains much under its capacity; it stood for a long time at Dundee, went to Aberdeen and hung about there, and did a very poor mileage per day, with the result that it showed a heavy coal consumption, most of the coal being burnt through misuse rather than in working trains."

 

Apologies if you think me too critical but it's a matter of interpretation. I highlighted your comment that Thompson was wrong because (a) evidence at the time was that the locos were not best suited to the Aberdeen route because of time spent not working although best suited in terms of working the heavier trains (b) the locomotive power would have been better used elsewhere where the demands justified the power (c) Thompson was working in a time frame that didn't know how long the war would last and (d) Thompson had no idea of what future demands would be made on the locomotives or the staff / loco resources available to meet them.

This is the point I believe SACM is trying to show in his proposed book. Did Thompson do the wrong thing by Gresley or the right thing for the times and situation then in vogue. History needs qualification and I understand that SACM is looking to judge on the context of the time.

 

I shouldn't have thought so. Having tried a bit of that sort of thing myself I bet that 90% of it is non contentious anyway.

Personally as someone who is no especial LNER or Gresley fan and who had barely heard of Thompson two years ago I still find the vitriol amazing. But it seems to me:

1) Thompson was greeted with reliability problems that needed rectifying. He attempted to rectify them
2) Times had changed. It was no longer the 1930s, and the business needs and priorities of the railway had changed too.
3) There were genuine and valid engineering reasons why Thompson wanted to undertake rebuilds and change policy, it wasn't just spite. (Whether those changes were the best possible is really a separate topic)
4) It seems irrational to make especial criticism of Thompson for failing to rectify issues that his predecessor and successor also failed to completely resolve.

And finally, were it possible to do a dispassionate analysis of the success of all the Big 4 CMEs at delivering what their companies required*, whilst from what I've read here I doubt Thompson would come out top of the list, at the very least there would be a dispute about who was at the bottom of the pile. There you are, there's a thread for you. "Rank the big 4 CMEs in order of their success at fulfilling their company's needs". That ought to generate enough heat to raise some steam...

*Something I think we enthusiasts have far too much baggage to be able to do rationally.

 

I am thinking of starting from scratch, or at least going over the original manuscript and changing it drastically.

My original views when I started this thread in 2012 have changed drastically and in the last six months so much more has come to light. For example, in response to some Raven centric views aired, I have been been researching Sir Vincent Raven like mad - when you start to find all of the evidence and strands regarding his relationship with his son in law, you can start to see the influence - and lack of it too - but it's a very human story of flawed men working very hard for their families and their employers.

The Stanier report is a big piece of the overall puzzle, as is the report regarding the crank axle failures on the P2s the P2 Trust recently found.

Overall I think I need to go further than Grafton, and above all reference my work as best as possible.

 

I'll be happy to respond when I get a moment from work and family life I'm afraid - my apologies for the sporadic responses. I've been very happy to sit back a little and watch the debate in addition to chiming in where appropriate. If you give me a few days I will try and respond as best I can.

 

one of the big problems with Thompson that applies to his rebuilds of the A1s and P2s (apart from the cost and drastic nature of the rebuilds) is that he completely failed to understand that with 3 sets of valve gear for a 3 cylinder engine you dont need to keep the conn rods the same length. from an engineering and design point of view he failed to understand the basics of valve gear design. had he grasped these fundamentals he could have produced a rebuild that was cheaper and sound and not suffer the frame problems that the rebuilt P2s and A1 had.

it appears that Peppercorn understood the above, but Thompson did not.

when Maunsell removed the Holcroft conjugated gear from those locos so fitted and substituted a 3rd set of valve gear, no major alterations were carried out to the locos at all. the alterations were economic and practical and sound. unlike Thompson's!

these comparisons (without the hindsight of the past 60/70 years) are important. i quite agree with Jimc on this point.

cheers,
julian

 

Julian - equal length connecting rods for all sets of valve gear used around the world and on the GWR. They can't all be wrong and Thompson is perfectly at liberty, like any CME, to plough his own furrow as it were.

Please stop quoting the rebuilds as "costly" - they were not any more costly (and were in fact considerably cheaper, as a result of using already existing components in many cases) than any prototype Gresley had previously built. In fact all of the rebuilds for prototypes or otherwise were, in the context of their use and in all cases bar exceptional circumstances (i.e. B3/3 and D Class aside) had long working lives with the LNER and then BR, only scrapped by the onset of steam - as with all steam engines of any designer.

The "frame problems" cited are interesting because there is quite a lot of evidence available that points to the Gresley A3s and Peppercorn A2s suffering worse cracked frames than any of the Thompson Pacifics (bar the A2/2s - which were spliced between the front frames and leading driving wheels, different in some details to the arrangement on subsequent classes). The A3s in particular had 3/4 frames replaced at varying times. I hardly think it is fair or representative to single out Thompson's Pacifics for this criticism when Gresley and Peppercorn locomotives usffered similarly or worse.

The A4s didn't suffer so badly and I can only assume it is down to the different in the frame stays and their placement. If anyone would like to enlighten me I'd be happy to listen. Bittern in preservation has also had partial frame renewal and one wonders if this is another criticism of Thompson that is somewhat overplayed - or given, ignoring the fairness of balance in reporting the other frame problems elsewhere.

 

Just a quick response here - the rebuilt P2s were returned to the Aberdeen line, and reports contemporary to their return there (and specific to black sheep Thane of Fife, which was the single chimney one and considered the worst of the six for various reasons) give surprisingly positive reports of the rebuilt prototype. No.995 - later 505 - had several reports in The Railway Magazine including a before and after article which was quite balanced. Certainly no hint of bias at this stage - it was a report concerned with the engineering involved.

The Peppercorn A2s were considered a better fit for Scottish routes and effectively replaced the older A2/2s, which were then sent further south (and this makes a lot of sense given the Eastern region tended to group similar classes together at specific depots).

 

Did these locos "around the world" also have divided drive and with the subsequent longer wheelbase as a result, a lower factor of adhesion? On quote I've seen from a driver regarding the Thompson A2s was "they'd slip on Mussleburgh Sands."
A prototype by nature will be costly due to the "one off" nature of the beast. To compare that to a rebuild is unfair. You need to compare the Thompson rebuilds to a Gresley rebuild of say a Raven B16.

 

One final thought. The major arguments for and against seem to centre on the debate between out and out technical performance and the ability to carry out duties on a daily basis. There's no doubting the P2s were capable of extraordinary and prodigious feats of haulage, but having an above average number of crank axle failures (and if the P2 Trust website is to be believed, this was a significantly under engineered component for the locomotive design) then you can understand the desire to eliminate crank axle failure and also removal of the conjugated valve gear.

In the report from Stanier, rebuilds of existing classes were suggested using three sets of valve gear. Thompson was already looking to develop a 6ft 2in "A Class" Pacific - why not use one of the P2s as a test bed in the vein suggested? It seems entirely reasonable on a number of levels.

He was designing for the future and for changing traffic conditions - one could reasonably argue he wasn't wrong to do this.

Firstly - yes. A number of locomotives around the world did have divided drive and subsequent longer wheelbases. Some of them even designed by Andre Chapelon, who converted a number of his own locomotives from one wheelbase type to another. Interesting parallels even if (and I happily accept this) Thompson is by no means in Chapelon's class of engineering.

On the second point, all Pacifics wheelslip due to the "sitting back" nature of setting off where the weight is transferred and the cartazzi or rear truck helps with this. That is a fact. Bulleid Pacifics have shown such a propensity themselves and I think it would be fair to say have shown some interesting characteristics in preservation from time to time where they are not handled carefully. They are perfectly when handled correctly however - never let it be said I am unfair.

One driver making a quote like that does not constitute a trend; however I think it is fair to say that their lower factor of adhesion did mean they had to be handled more carefully and there are several fair sources for this.

Regarding prototypes - some consistency, please. You (amongst others) can't keep moving the goalposts on this issue because it suits your views.

 

hi simon,

perhaps we are now getting to the nub of the problem. equal length conn rods are not required if each cylinder has its own valve gear. in this respect i am referring to inside -v- outside cylinders.

the GWR 4 cylinder locos had divided drive and in order to obtain full benefit from the balancing of a 4 cylinder loco the conn rods were the same length. plus the inside gear drove via cranked rockers the outside cylinders - and angularity of the conn rods if of different lengths (outside -v- inside) would have caused problems with the valve gear events.

that is a quite different situation to a 3 or 4 cylinder loco with 3 or 4 sets of valve gear.

these are quite fundamental basic design issues.

there are lots of examples of CMEs and their teams failing to understand valve gears fully. it appears that Thompson falls into this category. on the other hand the GWR and SR were very well served by geniuses Pearce and Holcroft. for a not very well known example, the SR Z class inside walschaerts valve gear is a very clever novel design and shows Holcroft at his most brilliant.

cheers,
julian

 

I know that but some are worse than others. Do you have the adhesion factor figures for the different Pacific classes?
I agree that one driver's comment does not make a trend but you used a similar argument regarding the Class D49 rebuild. I'd say that just because one driver liked it, it doesn't mean the rebuild was a success.
As for moving the goalposts - you're a dab hand at that yourself young man.
In the end you'll have a book to sell that exonerates ET and with that in mind, I can't see you ever agreeing to much criticism of the man.

 

I do as a matter of fact and the Bulleids do not come out very well I am afraid!

There is the extenuating factor that the driver in the case of the Morpeth was its regular driver for over two years and had reported back to his superiors on several aspects of the locomotive in question. So not quite the same there.

I agree.

Please - I'd like to think I've been entirely more flexible than you have been in your views.

Cynical, somewhat rude - but then your view is absolutely fixed that Thompson could do no right anyway.

Would you not agree that, if the evidence is there to support it, views can and should change? That if he deserves some exonerating, we should? Equally I think there are many things to be critical of with Thompson's reign as CME. I disagree fundamentally with the P2 rebuilding but that does not also mean I cannot look to understand his rationale behind it. The two are not mutually exclusive, and that I think is where your views show their entrenchment.

 

re the P2 crank axle failures, this isnt an argument for drastically rebuilding the class, but of beefing up the design and avoiding fatigue points on the key ways. something similar happened with MN Bibby Line in 1953 at crewkerne and new better crankaxles were fitted to the class.

i think Thompson is a bit like D. Earle Marsh of the LBSCR. Marsh produced one or two brilliant designs for which he is remembered, but also designed some real horrors, and generally it is thought the LBSCR would have been better served by someone else. Marsh also had a strong dislike for the old Stroudley locos, though many that Marsh designed to replace them werent up to the job and werent as good as the Stroudley locos. quite a few parallels here i think!

cheers,
julian

 

Why does wheelbase and divided drive affect the factor of adhesion? All else being equal, factor of adhesion just depends on TE (i.e. cylinder size, boiler pressure, wheel diameter etc) and the weight on the drivers. The specifics of the layout don't come into it. If you took a 2-8-2 and changed to a 4-6-2 but left everything else unchanged, it would presumably become more prone to slipping as some of the adhesion weight would be moved to the bogie - but that is not related to having divided drive.

Worth considering that in all the locos that I have calculated a factor of adhesion for (a non representative, but wide-ranging sample), on paper the most "slippery" should be the Adams T3!

Tom