What Ifs, and Locos that never were.

Hammer blow isn't actually a reason in favour of double singles, because coupling rods can be perfectly balanced and therefore don't have to cause any. Pistons and connecting rods, on the other hand, have to be balanced by other pistons and connecting rods, and a double single has twice as many. Unlike a conventional four-cylinder engine, where the cranks can be set to largely cancel this out, a double single can wander in and out of phase at will, and may therefore have no hammer blow, or the same amount as a pair of two cylinder engines concentrated into the position of one, or anything in between.

I've often felt Urie gets a slightly raw deal in the conventional memory of SR classes. His very successful developments of Drummond designs are remembered as Drummond engines, because they were rebuilds, while Maunsell's very successful developments of Urie's designs are remembered as Maunsell engines, because they were built new.

 

@8126 Last point first: I shouldn't be in the least bit surprised were @Jamessquared to make the same observation concerning Harry Wainwright's classes D, E & L !

Many thanks for clarifying 'the balance thing' though I understood the problem to be as much attempting to ameliorate the combination of the differing directions of the forces in play as anything else. They do say you learn something new every day!

I've been pleasantly surprised by how much information has survived to be so generously shared on these comparatively obscure and long extinct machines

 

Whilst that's true, it did occur to me that the 10ft coupling rods that would be required if they had been built as 440s would have had to be particularly heavy, and in turn require very large balance weights, which is all useless axle load that has to be subtracted from weight available for the boiler. Whether thats enough weight to be significant I don't know.
I also wonder if perhaps the two valves could have been driven off a single set of valve gear with rockers like the Churchward 4cylinder design.

 

It's the Q / Q1 that always amuses me: clearly Ashford saw the Bulleid loco as a Maunsell Q that had been reboilered, and in their normal style classified it as a Q1. But Bulleid gets the credit, whereas a Wainwright reboilering of a Stirling class O is still seen as a Stirling O1! (Perhaps if Brighton had been the centre of design for the SR, we'd talk about the Maunsell Q1 and the Bulleid Q1x ...)

Being serious, it does seem to show that as enthusiasts maybe we get too carried away on the cult of the designer as the single heroic figure, rather than seeing the role of the team, of which the CME is the figurehead but not the whole. The dynamic of how different design offices organised as teams, and the relative roles played within by the figurehead CME, is maybe an under documented area for study.

Tom

 

Tom, there's something in what you say, but it's a very Anglo-centric view of Drummond's career.
His early career with the Caley and the NBR produced locos which were some of the best around, and much emulated. Their descendants by other designers after his departure (but to his philosophy) served those railways commendably until the grouping.
Certainly later in life Drummond didn't seem to be able to move with the times, producing first designs which were just enlargements on a theme (but without proportionate performance) and then some rather ill-advised blind alleys. Incidentally his brother Peter did also in Scotland (Highland / GSWR).
On the LSWR there is the additional factor that he followed an incredibly able Loco Sup, one who for some reason is rather forgotten by history by who produced some of the finest locomotives of his day.
I don't think the failure to learn the lesson of Churchward is particularly distinctive: almost all other CMEs did the same. As an obvious example many others produced very disappointing four-cylinder 4-6-0s around the same time as Drummond.

 

Should we be looking at the chief/leading draughtsmen do you think?

 

I suspect that, if OVSB hadn't rather dropped the ball over the braking capacity of the Q1, that Ashford reasoning would've been somewhat harder to justify! In LBSC terms, surely Q would've become Q1 and what we know as Q1, Q2 .... or would it be .... no ..... wait a sec ....

 

I do apologise, while my explanation was technically correct, it was also terribly worded. To clarify:

Coupling rods are, like crankpins, a rotating mass in the locomotive's frame of reference. They can therefore be perfectly balanced with weights in the wheel rims, which are also rotating.
Pistons are a pure reciprocating mass in the horizontal-ish direction.
Connecting rods are a bit of both.

Any rotating mass can quite satisfactorily be balanced with more rotating masses. A reciprocating mass can either be balanced with more reciprocating masses, provided they are out of phase by more than 90 degrees and (preferably) sum to zero. Good combinations for this are two masses 180 degrees apart, or three at 120 degree intervals. Alternatively (as with most 2-cylinder engines) you can fudge it with more rotating masses. These can reduce the horizontal imbalance, at the cost of introducing a vertical reciprocating force, which makes your crew, goods and passengers feel better but which causes great anguish to the long-suffering civil engineer. If you over-do it the wheels lift off the rails at very high speed, which might be considered undesirable.

Drummond got to the answer later. The D15 4-4-0 manages to get an extra foot on the grate compared to the T9, with the same 10'0" rods, by carrying a sloping grate partly over the rear axle, like nearly all successful 4-6-0s and the chools class.

 

A little, but also perhaps think about the problem. I tend to agree with @andrewshimmin that Drummond's earlier designs tended to be a reasonably robust solution with some room for growth. So if a design team is asked for "One of those, but more powerful/stronger/faster," they'll tend to draw an improved version of the old design, if it will stand expansion, even if they didn't have much to do with it originally. Where a clean sheet is required, that's when you'll see the differences between design teams.

When Drummond left Scotland, his basic 4-4-0 design was what became the LSWR C8. This design stood up to two expansions on the LSWR, once to become the T9, very successfully, and again to become the L12, with diminishing returns. At that point the design concept could go no further and something new was required. I would guess that similar things happened in Scotland, even though Drummond and perhaps a lot of the leading figures of his Scottish team weren't there anymore.

 

If you mean were these designs more the draughtsman's than Drummond's, certainly I've never heard that suggested in any relevant books.
It's hard to square that with Drummond taking his philosophy with him from NBR to Caley (where he produced nearly identical locos, and he did again later on the LSWR: take a look at the Caley Jumbo and the LSW black motor, likewise the 4-4-0s and 0-4-4T), while at the same time the NBR was producing Drummond-derivative types themselves. Furthermore, brother Peter was producing almost identical locos on the Highland.

 

I suspect you then get back to the problem that, without connecting rods, the two (no longer)independent engines would still be trying to get in and out of phase, particularly if there were small differences wheel diameter, but be prevented due to the connection of the valve gears; which probably goes against the free-running reasons for trying the double singles in the first place.

That said, the use of different type valve gear on the inside and outside of the double singles seems fraught with issues and probably goes a long way to explain why at running speeds, the inside engine was found top be providing more than its share of the work. Surely it cannot have been beyond the wit of the drawing office to have come up with an arrangement of Stephenson's gear outside to give equivalent geometry to the inside motion?

Tom

 

What is interesting is that when Drummond came south, not only did he continue building locos that, as you say, were very similar to his Scottish locos; but that those locos were radically different from the Adams' designs that preceded them at Nine Elms. SO that suggests very particular ideas of his own and - who knows - maybe even access to drawings? Hard to see how designs could be worked up so quickly otherwise - though, in that light:

Another interesting point is that the LSWR at that point was still contracting out some loco building; and whereas under both Beattie and Adams Beyer-Peacock had been the generally preferred external builder, under Drummond the preference seemed to be the Scottish builders - Dubs or NBL.

Were there occasions when near identical Drummond-pattern locos were being built for LSWR and the Scottish companies side by side at the same external builders? I don't know enough about the build history of the Caley / Highland / NBR locos.

Tom

 

Apart from the addition of bogies and somewhat more generous steam passages, Drummond's designs seemed to owe a lot to Stroudley. For example, the marine big ends and various types of feed water heating. When, in later years, he departed further from the template, the troubles began.

 

Yes, in many ways Drummond's successful formula was to take the good elements of Stroudley's design philosophy, and marry them to more conventional wheel arrangements (i.e. 4-4-0 not 0-4-2 for express working), although he did also develop the type.

 

I don't know the answer, but I highly recommend the excellent book "The Scottish 4-4-0" by Tom Middlemass, which also looks at Stirling SER and Drummond LSWR types, and covers the Drummonds careers extensively.

 

Outside Stephenson's was definitely not liked. Interesting, for example, that when the GWR renewed the VOR narrow gauge 2-6-2Ts which had outside Stephenson's the updated design had Walschaerts' gear.

 

IIRC, the arrangement on the D&M 2-6-2t's was worse than described, being twixt wheels and outside frames. What fun for the staff. Wasn't there a comment from a senior GW man, along the lines of "what do you expect if you buy your locos from an injector manufacturer?"

 

A most interesting discussion for the last couple of pages on Drummond and his works. In one of your earlier posts, you offered the view that Drummond's greatest contribution to the LSWR was the re-location from Nine Elms to Eastleigh. That usually gets only a passing mention in the many books that talk about Drummond locomotives. Your post was a useful reminder that loco design is only one part of a CME's responsibilities. I cannot think of many railway workshop re-locations similar to that of the LSWR. The Sharp Stewart move from Manchester to Glasgow in the 1880s comes to mind, as does the Lancashire & Yorkshire move from Miles Platting to Horwich in the same decade, but the latter move was less than 20 miles.

Dugald Drummond's greatest design triumphs must surely have those from his initial period on the North British Rly - the Abbotsford class express 4-4-0 of 1876 and the corresponding 100-class and 34-class 0-6-0 goods engines. He cloned the Abbotsford and 100 classes for the Caledonian in the 1880s, and then did the same again on arriving at the LSWR, producing the C8-class 4-4-0 and 700-class 0-6-0 (the latter being a slight enlargement of the Caledonian "Jumbo"). But whereas the Abbotsford was in the forefront of British locomotive development in 1876, the C8 in 1898 fell short of some designs by then current on other railways, including the Dunalastair types that the Caledonian had been building for a couple of years. Drummond was able to evolve his standard express 4-4-0 through the T9, L12 and D15, but these engines achieved more success after Urie added superheaters, something that Drummond had resisted.

In his early days on the North British, Drummond had been heavily influenced by Stroudley but also showed himself open to other influences, notably taking forward Thomas Wheatley's early adoption of the classic British inside-frame, inside-cylinder 4-4-0. But I get the impression that, by the time that he reached the LSWR, Drummond was no longer interested in what other engineers did, merely in pursuing his established designs plus some idiosyncratic ideas of his own.

The "double single" T7 and E10 engines were notably unconventional. They had a De Glehn type cylinder layout, and some of the earliest De Glehn 4-cylinder compounds in the late 1880s also had an uncoupled "double single" layout. But the French quickly reinstated the coupling rods to produce conventional 4-4-0s. Webb on the LNWR stuck with uncoupled designs for 15 years, but even he abandoned the idea just as Drummond adopted it. It appears that Drummond wanted to get a very large firebox between the powered axles, but proportions got skewed - 27 sq ft grate was similar to that of the Schools and D49 4-4-0s, which had much larger boiler barrels and heating surfaces.

Peter Drummond usually followed closely in elder brother Dugald's design choices, but with a notable exception. When Drummond Junior took over on the Highland Railway, he inherited a design for a passenger 4-6-0 sketched out by his predecessor David Jones. Drummond accepted the design and added his own finishing touches to produce the Highland Castle-class, one of the most successful of the early British 4-6-0 types. If only Drummond Senior had taken note, the LSWR F13-class might have been a very different beast. Of course, a Highland Castle was designed for modest speeds on hilly routes and could not have handled expresses out of Waterloo, but it is tempting to think that it might have been useful in the West of England.

You mention Dugald Drummond's placement of orders with the Glasgow locomotive building firms. I believe that the Caledonian and North British built the majority of their engines within their own works at St Rollox and Cowlairs, but Neilsons and Dubs seemed to get most of the contract orders that were placed. Drummond very likely had maintained contacts in those firms, who would certainly have had many Caledonian and North British drawings and patterns already to hand. So when Dubs built a batch of Highland Railway "Small Ben" class 4-4-0s in 1898-9, which were a smaller-wheeled Abbotsford/C8 variant, they would have been dealing with a familiar type. The following year, it is recorded that Peter Drummond was able to negotiate a reduced price from Dubs for a batch of "Barney" 0-6-0s for the Highland. I wonder if the reduction had anything to do with the fact that a "Barney" was virtually identical with a Dugald Drummond 700-class 0-6-0, a batch of which had been built by Dubs in 1897? Dubs were at the same time building a batch of T9s for the LSWR, which likely shared many components with the Highland Railway "Small Bens".

While discussing affinities between engine types from different railways, the evolution of Wainwright SECR 4-4-0s does suggest some influence from John McIntosh and his Dunalastair evolution on the Caledonian. The SECR D-class and E-class are dimensionally very close to the Dunalsair II and III types, apart from the E-class adoption of the Belpaire firebox. The later SECR L-class is, Belpaire firebox again excepted, very similar both to the superheated version of the Dunalastair IV and to the equivalent North British Scott-class (LNER Class D30).

 

An excellent summary.
There is also another generic strain in this story of locos from Stroudley and Drummond in Scotland to the Southern.
S.W.Johnson (later of the Midland) originally worked with Drummond and Stroudley on the Edinburgh & Glasgow Railway, before both went off to the Highland (possibly to escape Johnson!).
Later Johnson's son James worked on the GNSR, where he designed a class of locos, a later version of which was sold by the cash-strapped GNST to the SECR in 1900.
I've always thought there is a strong visual affinity between Wainwright and S.W.Johnson (Midland) locos.

 

I've had a look in the Middlemass book, and the only batch of Drummond 4-4-0s built in Scotland was the T9s built by Dubs in 1899/1901.
About the same time they were building some of brother Peter's smaller wheeled but similar Wee Ben's (begun in 1898).
The NB and Caley classes were mostly built at Cowlairs / St Rollox. By this period Caley designs had progressed quite a bit to Dunalastair III level, quite a development from when Drummond left (Lambie and McIntosh being disciples of Drummond, but innovative in their own right too).
On the NBR the development was perhaps somewhat more Drummondish still, but even so Holmes was on his third generation of 4-4-0s already by this time, Drummond having left there twenty years before.