A Coarse Guide to the Steam Locomotive for 'N' Gauge Modellers - Part 88


Hello Chums

Compounding - A Simple Approach -

'Second Compound Era' from 1890 to 1922 - Comparisons:

Do you fancy playing a little game? Please think of five steam locomotive engineers - the first five that come to mind.

Lovely. Now please think of another five.

Super-duper. Am I correct in saying that George Whale is not on your list of ten?

By the way, please feel free to reply to this postington with your list - that would be fascinating to see.

The circumstances around Mr Webb's departure from the LNWR have been written about many times and I don't intend to divert our discussion by adding my own contribution. He retired in 1903 and settled in Bournemouth, dying there on 4 June 1906.

His successor as Chief Mechanical Engineer was George Whale. Mr Whale's recent background had been in the Running Department and he was especially aware of the shortcomings of Mr Webb's compounds. He took drastic action, comparable with what Sir William Stanier would do on the LMS thirty years later.

Amongst others, he introduced two important new classes, both conventional simple-expansion locomotives, the 'Precursor' class 4-4-0 and the 'Experiment' class 4-6-0. These were in the tradition  of Mr Webb's 'Improved Precedent' 2-4-0s and were straightforward engines capable of hard work. They were sorely needed. Crewe built 130 'Precursors' in three years between mid-1904 and mid-1907. One hundred and five 'Experiments' were built between 1905 and 1910.

These were especially capable engines and a relief to all concerned after years of struggling with the fickle compounds.  There was a mass withdrawal of the remaining three-cylinder compound passenger engines. The three and four-cylinder compound 0-8-0s were mostly rebuilt as two-cylinder engines and survived into the nineteen-sixties as class 'G2a'.

Mr Whale retired in 1909¹ and his successor, CJ Bowen Cooke, carried on with the conversions. Commencing in 1913, the 'Alfred the Greats' were rebuilt as two-cylinder simple-expansion locomotives called the 'Renown' class. As mentioned earlier, he also had the 'Bill Bailey' compound 4-6-0s withdrawn. A few compounds survived into the LMS period, the last being withdrawn in 1928.

So ended the fascinating story of LNWR compounds. A few other railways, notably the Lancashire & Yorkshire, had small numbers of compounds, but there was only one further large-scale use of the principle in Great Britain. Before we get on to that, let's have a look at a couple of instances where a fair comparison between compound and simple-expansion was made. As we have seen, Mr Webb's tests appear, and I mean no affront to his memory, to have been designed specifically to favour the compound.


*

HA Ivatt, Mr Ivatt the Elder, had commenced his apprenticeship on the LNWR, under Joseph Armstrong, in 1868. He joined the Great Southern & Western Railway in 1877, succeeding his friend, JAF Aspinall, as its Locomotive Engineer in 1886.

When in Ireland, Mr Ivatt kept in touch with Crewe - he was friendly with George Whale - and appears to have enjoyed a cordial relationship with Mr Webb. Perhaps it was this that gave him the idea to try compounding.

In 1895, he converted a 4-4-0 and an 0-6-0 to two-cylinder compounds on the Mallet/ von Borries/Worsdell system. Both were inside-cylinder engines and the Irish 5 ft 3 in gauge gave him rather more room than Mr Worsdell had on the North Eastern. As to the result, we are fortunate to have the words of his grandson:

'It amused Ivatt by showing neither gain nor loss from compounding.'²




[HA Ivatt's compound 4-4-0 for the GSWR]


Mr Ivatt had failed to find the 'something for nothing' pot of gold at the end of the compound rainbow. But he also avoided the problems that were apparent in Mr Webb's compounds - the early 1890s was the time of the idiosyncratic three-cylinder 2-2-2-2 classes.

In 1896, Mr Ivatt was appointed locomotive Superintendent of the Great Northern Railway - the English one. The railway's Board had made enquiries of several prominent locomotive engineers - including Mr Webb - before offering the appointment.

Mr Ivatt introduced the 4-4-2 'Atlantic' type to Great Britain in June 1898, beating his friend JAF Aspinall of the Lancashire & Yorkshire by a few months. No. 990 and her sisters were splendid locomotives. In 1902, the first of his large-boilered 'Atlantics', No. 251 emerged from Doncaster Works. These were truly excellent locomotives and the Great Northern was at the forefront of locomotive practice in Great Britain.

As amateur enthusiasts, we tend to overlook some of the constraints faced by locomotive engineers. We can all recite their names and achievements, but how many of us can do something similar with Directors or General Managers?

The Directors of the Great Northern, probably egged on by their capable General Manager, Oliver Bury, became interested in the possible economies offered by compounding. Mr Bury had a background in engineering and was fascinated by the potential benefits of compounding. Mr Ivatt advised against compounding but was overruled and the Board, on Mr Bury's recommendation, decided on 1 July 1904, to purchase a compound 'Atlantic' from the Vulcan Foundry Co.

Of course, prestige demanded that Mr Ivatt design his own compound 'Atlantic' and No. 292, a four-cylindered locomotive, emerged from Doncaster Works in March 1905.

The Vulcan Foundry engine, costing £4,000, also with four cylinders and resembling the compounds on the Nord in France, arrived in July 1905 but had some initial troubles and was only accepted by the GNR in October. It was numbered 1300 by the GNR.





Careful comparative trials between the two compounds and No. 294, a standard two-cylinder 'Large Atlantic', took place in 1906, and Mr Ivatt presented the results in his only formal paper to the Institution of Mechanical Engineers. HAV Bulleid provides full details in Master Builders of Steam³, so I'll only give the most important figures here:-

Cost (pence) per engine mile:

No. 1300, Vulcan compound: 3.125
No. 292, GNR Ivatt compound: 2.91
No. 294, GNR Ivatt simple: 2.88. 

Cost (pence) per ton-mile:

No. 1300, Vulcan compound: 0.0092
No. 292, GNR Ivatt compound: 0.0085
No. 294, GNR Ivatt simple: 0.0085.

All in all, a satisfactory result for Mr Ivatt. His compound used fractionally less coal than the simple 'Large Atlantic'. Interestingly, the Vulcan compound used the most coal.

The simple 'Atlantic' had noticeable savings in oil and repair costs and can be declared the winner by a small margin.

These trials did not take into account the higher initial cost of a four-cylinder compound compared to a two-cylinder simple which makes the standard 'Large Atlantic' a clear winner.

I believe the GNR trials of 1906 to be a fascinating test of the benefits which could be obtained by compounding in British conditions prior to the introduction of superheating.

However, Mr Ivatt was not completely convinced. He discussed the matter with Wilson Worsdell and George Hughes, and wondered if the LP cylinders on No. 292 were large enough. A second four-cylinder compound 'Atlantic', No. 1421, was built in 1907. She had the diameter of the LP cylinders increased from 16 in to 18 in. Unfortunately, Mr Ivatt didn't publish any details regarding No. 1421's performance in service. However, we can hear from HAV Bulleid again:

'No. 1421 did seem to have something extra, but not such as to justify the extra first cost and maintenance, Ivatt decided once again.'⁴





In the next part we'll have a look at comparative tests on the Great Western. 


¹ Unfortunately, Mr Whale didn't have a long retirement. He died a year later, in 1910, aged 67. I think it was entirely appropriate that Mr Bowen Cooke named the 4-4-0, No. 896, built in 1910, George Whale.

² HAV Bulleid, Master Builders of Steam, Ian Allan, London, 1963, Page 17.

³ HAV Bulleid, Page 32 et seq.

⁴ HAV Bulleid, Page 41.


'N' Gauge is Such Fun!

Many thanks and all best wishes.

Pip-pip

John








   

Churchward, Collett, Dean, Hawksworth, Gooch.

No surprises there then 

My second five

Drummond, Urie, Bullied, Adams, Stanier

No great surprise there either.

Gresley, Stanier, Bulleid, Collett, Churchward

followed by

Bowen-Cooke, Beames, Ivatt, Maunsell, Fowler.

Not sure if there are any surprises there or not.
PW

My "Top Ten" CMEs - in NO particular order - are G.J.Churchward, W.A.Stanier, O.V.S.Bulleid, H.N.Gresley, R.E.L.Maunsell, C.B.Collett, R.Riddles, W.Dean, R.W.Urie, and J.G.Robinson. Frankly, I found it impossible to place them in any order of merit or preference! I even have a "reserve" list of seven(!) more: D.Drummond, F.W.Hawksworth, H.S.Wainwright, G.Hughes, L.B.Billinton, H.A.Ivatt.

As anyone can see from the above listing, I am certainly no engineer, having a Degree in Graphic Design/Visual Communication. My engineering knowledge consists of two principles:- 1) If it moves - grease it; 2) If it's meant to be fixed - knock it in harder!

In my opinion this "mini" series of articles is utterly brilliant, and really ought to be considered for turning into a book!

I'm now looking forward to the next episode...

Quote from: port perran on April 16, 2025, 03:05:06 PMMy second five

Drummond, Urie, Bullied, Adams, Stanier

Ah, but which Drummond?   

BTW, it's Bulleid 

A Coarse Guide to the Steam Locomotive for 'N' Gauge Modellers - Part 89


Hello Chums

Compounding - A Simple Approach -

'Second Compound Era' from 1890 to 1922 - More Comparisons:

Let's begin with a questionton. What is the connection between the Webb three-cylinder compound 'double-singles' on the LNWR and GWR 'Star', 'Castle' and 'King' Classes? Or, if you are of the crimson lake persuasion, the LMS 'Princess Royals'.

**

'de Glehn' - synonymous with French steam, especially compound locomotives. A Frenchman? No and yes. Not by birth, but by adoption.

Alfred George de Glehn was born in SuperSunny Sydenham, in 1848, 'd'un père d'origine balte et d'une mère ecossaise.'¹ After attending Kings College, London and Zürich Polytechnic, he joined Société Alsacienne des Constructions Mécaniques (S.A.C.M.)at Mulhouse. His progress there was rapid and he was head of the firm by 1878.

S.A.C.M. was closely connected with the Chemin de fer du Nord (Nord) of France, which had adopted for its express passenger engine a modified version of the 2-4-0s built by Archibald Sturrock for the Great Northern Railway. The type was enlarged over time and M. de Glehn suggested fitting a front bogie to a batch S.A.C.M. was building for Nord. These locomotives became known as the 'Outrance' class and 50 were built from 1879 to 1882.




[A GNR Sturrock 2-4-0. The Nord locomotives were similar even having inside cylinders and Stephenson link motion.]


Nord approached S.A.C.M in 1884 about building another 12 'Outrances' and M. du Bousquet, its Chief Mechanical Engineer, mentioned there had been some cases of frame fracture.

M. de Glehn thought about the stresses that were causing the fractures and arrived at a proposed solution. S.A.C.M. would build, at its own risk, a prototype of a new locomotive type for the Nord.

Are you ready for this? The inspiration for this new type was Mr Webb's three-cylinder 2-2-2-0 compounds on the LNWR. The de Glehn locomotives had four cylinders, LP outside, set well back, driving the trailing axle and HP inside, driving the leading axle. There were four sets of Walschaerts valve gear. Nord No. 701, of 1885, was the world's first de Glehn compound locomotive.  Most wonderfully, she survives to this day.

No. 701 was reasonably successful and many more engines de Glehn compounds were built. These had the HP cylinders outside, with the LP ones inside, and had coupling rods - 4-4-0s. They retained the divided drive and the outside cylinders continued to be set back. These were the first of Les Chocolats and were eventually succeeded on the Nord by de Glehn-du Bousquet 'Atlantics' and then 'Pacifics'. Although the first de Glehn compounds had a low boiler pressure of 156 psi, the later engines had boilers pressed to 227 psi. 

Therein is the paradox - the de Glehn compounds became famous for fuel economy but, initially, the idea was to minimise mechanical stresses and consequent damage.

**

Now to the Great Western Railway. As early as 1901, before his appointment as Chief Mechanical Engineer of the GWR, Mr Churchward had drawn up outline diagrams for six new locomotive types, one of which was an express passenger 4-6-0 with 6 ft 8 1/2 in diameter coupled wheels. All types had two outside cylinders with inside Stephenson link motion, much influenced by contemporary US practice. Once Mr Churchward had replaced Mr Dean as Chief Mechanical Engineer, the scene was set for the emergence of 4-6-0 No. 98 in March 1903. She can be considered the first of the 'Saint' class.

However, Mr Churchward was not completely convinced this was the correct approach for an express passenger engine and obtained Board approval to purchase a de Glehn compound 4-4-2 from S.A.C.M.. No. 102, La France, was delivered in October 1903.




[The first GWR de Glehn compound 'Atlantic', No. 102 La France.]


The second 'Saint', No. 171 Albion, entered service soon after, with an important change. The original standard boiler pressure of 200 psi had been increased to 225 psi to correspond with that of La France. In order to ensure a fair comparison, Albion was converted from a 4-6-0 to a 4-4-2 in October 1904.

During 1905, another two French compound 'Atlantics', to a larger design, were purchased. And 13 more two-cylinder 4-4-2s were built at Swindon.

Clearly there was a great deal of experimentation and trial running but, as far as I'm aware, no evidence of the test results survive. However, Mr Churchward was clearly impressed by the smooth running of the four-cylinder compounds and designed a four-cylinder simple-expansion engine, along similar lines, to establish whether he wished to incorporate compounding in future locomotives .

This was No. 40, North Star, which appeared in April 1906. She must be regarded as one of the most important British steam locomotives ever built. As a 4-4-2 she suffered from having a lower adhesive weight than a 4-6-0 and North Star proved too powerful for her adhesion, which led to bent coupling rods. Mr Churchward was also mindful that a 4-6-0 would cope better with the Devon banks than a 4-4-2 and decided to make the 4-6-0 his standard type for express passenger work.

The 'Saints' that were 4-4-2s were rebuilt as 4-6-0s, as was North Star in 1909.  By then, 30 production four-cylinder 'Star' class 4-6-0s had been built, the first in February 1907. These were all simple-expansion, Mr Churchward having satisfied himself that the additional complication of compounding was unnecessary. The 'Stars' had two sets of inside Walschaerts valve gear.




[Mr Churchward first four-cylinder locomotive, No. 40, when still a 4-4-2. Later named North Star, converted to a 4-6-0 in 1909 and renumbered as No. 4000. She was rebuilt as a 'Castle' in November 1929 and withdrawn in May 1957. Her total mileage was 2,110,396. She was the only 'Castle' to run two million miles although No. 111, Viscount Churchill, rebuilt in 1924 from 4-6-2 The Great Bear, came close.]

There is, I believe, another factor we need to consider. Mr Churchward was the first British locomotive engineer to fit a superheater, to 'Saint' 4-6-0 No 2901 in May 1906. Over the next three years he experimented and the Standard Type 3 superheater was perfected in 1909. From 1910, all new GWR main line tender locomotives were provided with superheaters.

Superheating provided greater economy than compounding with less complication. Accepting, if we may, the proposition advanced in the the first part of this section that, in engineering, there isn't 'something for nothing', it becomes a matter of whether the benefit is worth the cost.

Returning to our opening analogy, imagine you deciding to forego compound interest so that we could, every year, each have a pint of fine, foaming ale and a lovely smile and cheery wave from the super big gel behind the bar in my local Wetherspoon's. Would the benefit of your compound interest be worth the cost of fine, foaming ale, lovely smile and cheery wave foregone? No!

But, imagine the Wetherspoon's closed and all we could get for your fiver is two halves of cooking lager served by a surly git of a barman with an attitude problem. Would the benefit be worth the cost. Yes!

It was decided quickly, albeit not quickly enough by some locomotive engineers, that the benefit of superheating was worth the cost.

RIP compounding? Not quite - there's a bit more to the story and, in the next part, we'll start off in Gateshead.

**

Let's conclude with the answer to the opening questionton.

The four-cylinder simple-expansion GWR 4-6-0s owed much to the de Glehn compound 'Atlantics'. 225 psi boiler pressure, Walschaerts valve gear, albeit only two sets on the GWR, the excellent de Glehn bogie and divided drive with the outside cylinders set well back - a feature M. de Glehn derived from Mr Webb's LNWR three-cylinder compounds!

And, of course, Sir William Stanier's 'Princess Royal' 4-6-2s for the LMS incorporated these features derived from M. de Glehn's practice. And like the de Glehn four-cylinder engines, this class had four sets of Walschaerts valve gear.

**

Coda

From Harold Holcroft's Locomotive Adventure:

'After the [Star] class had been in service for some months, the locomotive inspectors were able to express the opinion that the four-cylinder engines were a 'coach better' than the 'Saint' class on fast trains of 12 to 14 coaches, though both classes carried the No. 1 Standard Boiler. The superiority was put down to smoother riding and to the shorter cut-off possible with the Walschaerts gear.
The two-cylinder engines were the better in getting away after a stop...'²

Between 1906 and 1914, 'Stars' and 'Saints' were built, usually 10 each year. They sort of alternated, not exactly by year, but more generally. So the 'Saints' must have had something. Their better acceleration would be due, at least in part, to the variable lead of the Stephenson link motion having a reduced lead when starting and at low speed. As the engine was 'linked-up', the lead increased. The 'Star's' Walschaerts valve gear had fixed lead which would have been a slight disadvantage at long cut-offs.

The 'Stars' had an extra inch of valve travel.

But is this the whole story? Was 'a coach better' due to better riding and Walschaerts valve gear with a longer travel? I wondered if there was more to it.

Mr Holcroft was a very fine locomotive engineer and the inspectors would have been especially experienced enginemen. Surely, if there was another factor, Mr Holcroft would have mentioned it in his book.

Then I decided to look into the obvious - cylinder volume.

'Saint' two cylinders: 18 1/2 in x 30 in.

'Star' four cylinders: 15 in x 26 in.

Perhaps any GWR aficionados would kindly confirm these dimensions are correct.

Now, as you will have established, I'm a State-Registered Half-Wit and I'm hopeless at sums. I'm a 'one, two, three, many' sort of chap. So please take what follows with a wheelbarrow full of salt.

Cylinder volume 'Star' = 18,378 in³.

Cylinder volume 'Saint' = 16,128 in³.

Fascinating.

I know there are lots 'N' lots of brainy +++ chums on our FabulousForum. I'd be grateful if someone would check my sums.


¹ C Lecomte, La Vie du Rail, September 26, 1965.

² Harold Holcroft. Locomotive Adventure, Vol 1, Ian Allan, London, Page 58.


'N' Gauge is Such Fun!

Many thanks for looking and all best wishes.

Cheerie-Bye

John         

Quote from: Train Waiting on April 18, 2025, 10:22:17 AMCylinder volume 'Star' = 18,378 in³.

Cylinder volume 'Saint' = 16,128 in³.

Well John,

I would say that's nigh enuf fer pitwark!

Just depends  what flavour of pie you decide to bring to the party 

Cheers,

Colin.

Quote from: cmason on April 18, 2025, 03:07:13 PMJust depends  what flavour of pie you decide to bring to the party 

How about 95 days time?

    

Quote from: chrism on April 18, 2025, 03:22:15 PMHow about 95 days time?

    

@chrism        - good one.

Took me few seconds to get it since was already after breakfast Saturday here by the time I caught up messages.

Quote from: cmason on April 19, 2025, 12:33:28 AM

Quote from: chrism on April 18, 2025, 03:22:15 PMHow about 95 days time?

    

@chrism      - good one.

Took me few seconds to get it since was already after breakfast Saturday here by the time I caught up messages.

Yes, unfortunately with a site that has members from all around the world it's one that doesn't last long 

Different date formats in some parts doesn't help either.

I still don't get it!

It must be an "in joke" amongst the northern hemisphere folk as I don't get it either.

Quote from: Bealman on April 19, 2025, 06:26:55 AMI still don't get it!

Quote from: Graham on April 19, 2025, 06:48:31 AMIt must be an "in joke" amongst the northern hemisphere folk as I don't get it either.

An approximation for the value of Pi is 22/7.
95 days from when I originally posted it is 22nd July, sometimes jokingly referred to by mathematicians as "Pi Day"

Aha, very good. You learn something every day. Do not let anyone say this forum is not a treasure trove of knowledge!

Actually, that reminds me... I'm having a pie for dinner tonight