The rise of the 4-6-0 and why were there none on the GNR

[drmditch]

I have recently been reading ES Cox's 'Locomotive Panorama' Vol1, and Hamilton Ellis' History of the Midland Railway (thanks to the Shed Shop at NYMR Grosmont), Not being able to get to sleep last night (for unrelated reasons and nothing to do with worrying about 4F axleboxes) I found myself speculating on the use and development of what became the 'classic' 4-6-0 locomotive.

A quick analysis indicates that of the 16 major pre-grouping companies (apologies to any fervent disciples of the railways I may have left out):-

6 (SECR/LBSC/Midland/GN/NB/GNS) made no use of 4-6-0s

10 used 4-6-0s in some form or other, and of those 10,
4 (L&Y ??/Caledonian ??) GER and the G&SW) used 4-6-0s for express passenger haulage only, and
1 (NER - by grouping anyway) for 'mixed traffic' only. The remainder:-
5 (LSWR/GW - of course/LNWR/Highland/GCR) used this wheel layout for a range of traffics

One might suggest that the Churchward and Urie locomotives became the 'generic' types from which most of the later ones were developed (with apologies to Raven and Smith (of the Highland river class)).

Of the six without 4-6-0s, three (GN/SECR/LBSC ???) made use of 2-6-0s

In the case of the GN the largest 2-6-0 clearly overlapped in power and competence in traffic some of the 4-6-0s, especially the B7 and B16.

What were the reasons therefore why HN Gresley - a noted proponent of large locomotives - avoided the 4-6-0?

One might have thought that the O2 could have provided a 4-6-0 sister design with relatively little difficulty. One can think of other examples like O5/B6 and Q6/B15, even eventually on the LMS the Standard 5 and the 8F.

Was it :-
Cost of extending turntables?
Cost of a bogie as against a pony truck?
Less adhesion weight than a 2-6-0?
'If a 2-6-0 will do the job' why change?

The issue of three cylinders vs two would have been no more an issue than it was with 2-6-0s anyway.
Obviously a 4-6-0 would not permit a wide firebox, but the 2-6-0s wouldn't either.

Of course B7/B16/and B12s were all built to pre-grouping designs and improved somewhat under the LNER.
But, if a Fowler compound 4-6-0 might have looked a likely development for the LMS , why did 4-6-0 development on the LNER cease with the grouping?

PS - for the moment ignoring the GE local pressures that lead to the B17 - and might it not have been cheaper to rebuild the structures that precluded A1s/V2s) - but that is another whole question.

PPS - and not forgetting of course the eventual creation of the beautiful B1 in 1941 !!

[John Palmer]

Speculating with very little knowledge of such matters, could it be that HNG wished to avoid perpetuating the poor riding qualities of his moguls - K2 and K3 - by adoption of designs making use of a trailing truck? Another factor might have been a desire to extend the use of wide grates where possible, possibly with an eye to increased use of mediocre steam coal?

I also wonder whether the successful employment of a trailing wheelset both in the GNR's Atlantic and Pacific designs may have contributed to the formation of an institutional bias towards such a design feature.

[john coffin]

surely it is more important to consider the circumstances after the appointment of HNG?

3 years away from war, and already the government and many industries were working on war work. the railways were looking at improving
their reliability, and the GNR was still expanding its freight business, so the demand was for more freight locos, initially J6, then the 2-6-0's
but notice that he produced only a few K1/K2 but over 50 of the 6.

Indeed, before the First War, there was no real need for passenger locos, the Atlantics had not been found wanting in their original form,
that did not happen until quite late in the war when the carriages became dramatically overloaded, and the railway very run down.
By the end of the war, the H4/K3 were proving that a big boiler loco was the way forward with passenger stock, but still with the problem
around the size of the firebox without a rear carrying axle.

Realistically only the GWR solved this problem, I mean Robinson on the GCR created almost 10 small classes of 4-6-0, and none seemed to
have been that succesful, particularly in terms of steaming.

By the end of the war, Gresley had already worked out some plans for Pacifics, and seen the K4 details from America, and perhaps more
importantly was designing locos for a main line, not for potential use on lesser lines. Many forget that the original main purpose of the GNR was
to move passengers and freight to Scotland, and any powerful loco would need to do that rather than be a compromise like some would
consider a 4-6-0 to be. Remember that the Raven Pacific was only an extension of his 4-6-0's and may well not have been that successful either.
I know the tests were designed to prove the Gresley Pacifics, but from what I have read, the Ravens were improved with Gresley boilers.

The final important thing is MONEY, the GNR like all railways was quite poor after the war, due to the lack of recompense from the Railway
Board which had run them during the war, so had to be circumspect when spending and very soon the Grouping Act was in view, hence
the need to consider those items which were liable to influence the direction of the new Grouped Railway.
Paul

[kudu]

Drmditch makes a good summary of the 4-6-0 position at the Grouping. I would only add the statistics that over half (just) the 4-6-0s were on the LNW, where, uniquely, they were the most common type. On the other hand, the Highland had the highest % of 4-6-0s, even though slightly outnumbered by their 4-4-0s.

I would suggest that 2-6-0s might not be the best and certainly not the only alternative to consider, especially in the GN context. For express work the obvious alternative is the Atlantic, used on 7 railways - or 6 if we exclude the special case of the 3 de Glehn compounds on the GW.

In summary (companies listed in order of total loco stock):
The LNWR, GWR (ignoring the de Glehns), GER, Caledonian, LSWR, GSWR and Highland used 4-6-0s only.
The GNR, NBR and LBSCR used 4-4-2s only.
The NER, L&YR and GCR used both.
The Midland and SECR used neither, sticking with 4-4-0s (as did the Furness, GNSR and Cambrian, to take the only other companies that might conceivably have used 4-6-0s).

From this perspective the main questions to emerge are why the Midland in particular used neither and why 3 companies used both 4-6-0s and 4-4-2s. As for the GNR, I had always understood that the preference for a wide firebox was the main consideration, hence the choice of Atlantics. (No express 4-6-0s were running when the first GN Atlantics were designed.) Later maybe Gresley had also observed that other companies' 4-6-0s did not necessarily live up to expectations, especially for speed, as distinct from power. This seems to have been the case on the NER and, especially, the GCR, which seemed to lack direction in the variety of small 4-6-0s produced, each class in small numbers, while even the LNWR and LSWR express 4-6-0s were coal guzzlers (if Gresley knew that).

Concerning the NBR and LBSCR Atlantics, this site reasonably suggests 4 driving wheels were to be preferred for the Aberdeen route (I know Gresley had different ideas later) but I'm not aware of the factors behind the choice for the Newhaven route, where I guess the boat trains would have been relatively heavy.

Of the 3 companies that used both types, the NER and GCR both tried 4-6-0s before introducing 4-4-2s but on the L&Y it was the other way round. Maybe experience with Atlantics led the L&Y to go for the extra adhesion of a 4-6-0 given the frequent stops on their express route, but I'm guessing.

In LNER days Pacifics were preferred to the 4-6-0s used on the other 3 railways (the LMS came late to the Pacific party) and when a slightly smaller version was needed of course we got 2-6-2s before 4-6-0s.

I hope I have given a balanced account and no doubt will be corrected if I haven't.

Kudu

[Hatfield Shed]

...I also wonder whether the successful employment of a trailing wheelset both in the GNR's Atlantic and Pacific designs may have contributed to the formation of an institutional bias towards such a design feature.

Start from Stirling's 1870 4-2-2 eight foot single. Doncaster's achievement in retention of knowledge and design continuity from this loco to the end of its independent existence as a design shop is unmatched in the UK.

The progression 4-2-2, 4-4-2, 4-6-2, for maximum power express types is unique. Each provided a good vehicle with characteristics suitable for development by progressive enlargements and refinements which are well documented. Credit is not often given to Ivatt and Gresley for their both having 'picked up' on the Doncaster know how as the basis of their designs, rather than insisting on their own ideas. That of course the specialty of a later CME ...

An overlooked aspect in design appraisals between the 4-6-0 and 4-6-2 is the frame strength problem for the pacific. The UK railway 'stuck' at 20 and then 22 ton axleload maxima in the steam period post WWI ; enough for a 4-6-0, but marginal for a pacific with the capability to put much larger sustained power outputs through a frame of the same structure weight and strength as the 4-6-0. (This almost certainly the reason for Crewe's adherence to the 4-6-0: the cylinders of their pacifics came loose with great regularity - both types.) Doncaster got there in the end, and Bulleid was the beneficiary of this know how, which happily then made its way into the Riddles designs much to the benefit of all these classes.

[john coffin]

Hatfield Shed is correct to re state the continuation of design features at Doncaster both for Locos and Carriages from the start of loco building
in 1867. However, I reiterate that the fact that the GNR was predominately a main line railway, whereas both of its partners on the route to Scotland
had myriad branches for instance York to Harrogate or Scarborough which demanded a more eclectic range of locos. It is interesting that the
NBR did not follow its Scottish neighbours by going to a 4-6-0 sticking instead with the 4-4-2 as largest passenger loco. Also one wonders why
Raven went through all types, ie 4-4-2/4-6-0/4-6-2.

As for the GNR, it did not need many other passenger types until the Atlantics had been completely outclassed on the Manchester and Sheffield
routes, hence HNG having to wait until 1921 to start producing the A1's, which was also a product of lack of finance post First World War.

Paul

[Hatfield Shed]

Very true of the GNR being a fast main line between London and York(shire) in primary conception, and always the priority afterwards despite additions.
It was observed by one commentator in the 1900s that in locomotives the GNR reserved all its sophistication for the express machines, all the rest simply functional, strong and pretty basic to the point of a little rough around the edges. That would change after Gresley's appointment of course.

Without wishing to denigrate the other major LNER constituent's loco engineering practise, the maximum power express loco development seen at Doncaster looks very logical and ordered compared to that of the GCR and NER in particular, where it looks more like 'try everything and anything!' - Robinson it must be said seemingly had no clue, although the exterior appearance was often rather fine. The GER and NBR more ordered, progressing through enlargements of the 4-4-0: the GER to a 4-6-0 which is recognisably a big 4-4-0 and very good too, true horses for courses to comply with the GER's infrastructure restrictions; while the NBR's very successful 4-4-0s are abandoned for a narrow firebox atlantic, which as elsewhere could just as well have been built as a 4-6-0 for the adhesion advantage, and might have lasted longer in service if built to that format.

Compare to the freight loco development, which is rational and ordered at all these locations. Steady progress was possible by enlarging on past practise, with general success attending. Whether the terminus was a large 0-6-0, an 0-8-0 or 2-8-0, these designs would be maintained and as required developed to improve performance, only being withdrawn after fully economic working lives.

The LNER truly 'got lucky' twice over in the traction department: in inheriting Doncaster's experience, with a CME in the chair who knew how to use it for fast mixed traffic and passenger designs; and altogether a good fleet of well designed locos for the main revenue earner, freight traffic. The latter probably saved the LNER from yet more severe financial difficulty in the 30s, the scrap and build to replace inadequate traction that proved necessary on the LMS fortunately not required.

[john coffin]

Well said, it is interesting to look carefully at the problems of the LMS until they appointed Stanier where each of the works were fighting with
each other, and who could reasonably have thought that the evolution of the Midland Compound would be more logical than developing the
LNWR or even L&Y big engines?

We think we know that Gresley was not necessarily first choice for CME of the LNER, seemingly both Raven and Robinson were considered first, but being somewhat older passed the job over. Yet Robinson outlived Gresley, the pressures of command methinks.

The Board of the LNER were controlled at lest initially by a strong man who put the fate of the whole company before that of individual parts thereof, so it seems to have been decided quite early that only one works would prevail in terms of design and so on, and as you say, fortunately Doncaster was chosen. Looking carefully at for instance the NER even in 1922, the three works were still fighting with each other, hence it made economic sense
to look for a works that had a central role and power, and fortunately, that was where Gresley was so there was less in fighting than at other railways, although ET seemed to plough his own furrow at Stratford for some while.

The Goods locos of the GNR were certainly a careful evolution and one is constantly surprised by trains that were pulled at quite decent speeds by
even the most basic of them. Even the J50 was very much an evolution of the saddle tanks that went back to the beginning of the railway.

Paul

[4812]

Three minor points:
1. The CR did build mixed traffic 4-6-0s, two classes with 5'9" DW and one with 5'0". The ex-HR 'Rivers' gave another.
2. In addition to the French compounds, the GWR did try both 2- and 4-cylinder Atlantics, but found them wanting and converted them to 4-6-0s.
3. The GCR built pairs of near-identical 4-4-2s and 4-6-0s. The company multiplied the Atlantics but chose not to convert the 4-6-0s.

[earlswood nob]

Good morning all

This thread is a very interesting read.

Although it's not GNR, Gresley did own a proposed 4-6-0 for the LNER. It is illustrated in FAS Brown's biography of Nigel Gresley. It had a 6'0 boiler tapering to 5'9, three cylinders driving the middle axle and 6'8 driving wheels.

It was designed around the same time as the V2, and I guess the V2 performed all the duties envisaged for the 4-6-0.

Earlswood nob

[earlswood nob]

Further to the remarks on Scottish mixed traffic 4-6-0s.

Caledonian McIntosh 55 class with 5'0 driving wheels for the Oban line.

Caledonian Pickersgill 191 class with 5'6 driving wheels for the Oban line.

Caledonian McIntosh 908 & 179 classes with 5'9 driving wheels for mixed traffic duties.

Earlswood nob

[Hatfield Shed]

Although it's not GNR, Gresley did own a proposed 4-6-0 for the LNER. It is illustrated in FAS Brown's biography of Nigel Gresley. It had a 6'0 boiler tapering to 5'9, three cylinders driving the middle axle and 6'8 driving wheels.

It was designed around the same time as the V2, and I guess the V2 performed all the duties envisaged for the 4-6-0.

Very close in outline capacity to Crewe's final 2A boiler 4-6-0 group of rebuilt Royal Scot/Patriot/Jubilee; and I suspect much the same in performance - for good and ill. You get the good power performance of the proven engine section while the boiler steams well, and this from a lighter locomotive, so it all looks very impressive as it initially matches pretty closely what the pacific delivers in normal service.

But after circa 200 miles express on 400 or more tons, it's all done, the ashpan is choking off the air supply to the grate and power production falls off. Thus the classic operational image of the LMS and LMR: two 4-6-0's on express, one of them with mucho black smoke at the chimney. Had the Gresley big 4-6-0 been constructed, the K3s, B17s, D49s and B1s would have been piloting them all over the place.

This was the real unseen benefit of the wide grate over a carrying wheel: the greatly enlarged ashpan volume enabled the whole ten ton pacific tender load of coal to be burned, and still the airflow to the grate was unimpeded. Notice that Swindon's express 4-6-0s only had 5 tons of coal as designed, and that was good for their longest runs because on 'the billiard table' section relatively small power production was required - less coal burned - and combined with the use of a very high grade coal with little non-combustible solid content this all added up to less ash as a result. In 1925 it was observed that the Castle was 'all done' after London - Leeds, because the ashpan was full: a big Gresley 4-6-0 would have performed no differently, it's a fundamental limitation of the narrow firebox layout.

It's only the UK - of the major centres of steam development - that there is this obsession with the 4-6-0 to the point of a near religious belief in its powers; see 5AT project. France, Germany, USA and Canada, etc. etc. all move to the wide firebox for the advantages it brings when sustained power production is required. Ivatt and Gresley were ahead of the game in the UK, in early finding the optimum path to greater power output from the UK steam loco and being prepared to stick at it despite the development problems that came with it.

[earlswood nob]

G'day all

Thanks Hatfield Shed and others.

I am learning a lot from this thread.

Earlswood nob

[Wainwright]

Regarding the comments about the SECR and 2-6-0s, in the years before WW1 and the grouping the SECR under Maunsell (having seen multiple attempts by Wainwright to get either a 4-4-2 or 4-6-09 off the ground fail due to under-investment in track) embarked on a policy focused on 3 (and a half) main points:

- Finish building Wainwright's 'L' passenger 4-4-0s since they were already committed to do so. Wainwright's assistant/power-behind-the-throne Surtees, was just retiring and he managed most of the project.

- Build a new mixed-traffic 2-6-0 based on the GWR experience (at the point when first Pearson and later Holcroft, 2 of Maunsell's 3 key staff, joined the SECR, the GW had not yet built the Halls and its only mixed traffic 6-coupled locos were the 43xxs; Churchward was still seeing a mixture of 2-6-0s and 2-8-0s as his mixed traffic solution)

- Simultaneously they committed to ensure that this class had key parts in common with a new secondary passenger tank loco (the 'K' 2-6-4Ts) in essence an expanded Large Prairie borrowing from the Midland experience with 0-6-4Ts, the ex-Tilbury 4-6-4Ts and some never-completed Deeley 2-6-4T plans (Maunsell's draughtsman, Clayton, his other key collaborator and junior to Pearson but senior to Holcroft, was a former Midland man). In the event I think the 2-6-4T design was finished first, but the 2-6-0s appeared first.

- When the 'L's proved inadequate they rebuilt earlier Wainwright classes (The Ds and Es into the D1s and E1s - with much more Clayton involvement than the others, and so much more Midland influence) - the money was just not there for more new-build express passenger locos as well as new-build mixed-traffic locos.

Later, when under the Southern the 2-6-4Ts (renamed 'Rivers') seemed to have balance difficulties, they were rebuilt as the only passenger 2-6-0s in the UK ever, and built some more because they liked them (including some 3-cylinder ones, an experiment extended due to very tight loading gauges on some lines), thus driving down the demand for new 4-6-0s.

So, perversely, it could be argued that it was the GWR experience - the supposed '4-6-0 experts' - that led the SECR to build 2-6-0s, not 4-6-0s.

The exchange of ideas between railways was in some ways an odd process, often reactive and not always as logical as we would think with hindsight.

This is all in Holcroft's autobiography. Don't trust Wikipedia on this.

[Wainwright]

Oh, and the LBSCR was building very large 4-6-4Ts for fast passenger service right before grouping, as was the GSWR (which had previously built both 4-6-0s and 4-4-0s). People often write off the 'large tank' experiments that went on in this period on the shorter railways, because many of them were later scrapped.