The Background To Bulleid's Revolutionary 'Merchant Navy' Class

Précis of an illustrated talk given by contributor David Smith.    

Following the Railway Grouping of January 1923, the newly formed 'Southern Railway' appointed as its first Chief Mechanical Engineer, Richard Maunsell, who had previously held the post at the S.E.C.R. for some years.  His first task was to assess the loco stock from the former constituents, and develop a series of standard designs that would be suitable across the whole of the new 'Southern'.

Moving out of war time depression, the 1920s saw passenger traffic increasing, revealing a lack of suitable locos to haul heavy Expresses, particularly boat trains to Dover and Southampton, and fast West Of England services; instead relying on double or even triple-heading !

Maunsell's own SECR design 'N' Class Mogul had already proved to be a good all-round workhorse, but was clearly not going to haul a boat train on its own.  Maunsell also had the prototype 'River' class tank engine.  This was the loco that he had designed for Express passenger use on the SECR - it was similar to the 'N' class Moguls but had larger driving wheels at 6' compared to 5'6".  However, in 1923 only the prototype existed, other class members being delayed due to a backlog of war time repair work.  But anyway, being a tank engine, its range was limited and was not going to be suitable across the whole of the Southern.

                                                  
By 1926 Maunsell had put the finishing touches to his brand new loco, which emerged from Eastleigh works.  This was the 'Lord Nelson' class named after famous British Admirals, a four cylinder design with 4 sets of valve gear.  Theoretically it was the most powerful loco in Britain when built, although it soon lost this crown when the GWR built the King class in 1927.  Maunsell was cautious with this design, and wanted to test 'Lord Nelson' extensively before building any further members, so the loco was worked on heavy South West expresses.

Whilst this was happening, an incident occurred which was in part to have some bearing on the design of the Bulleid Pacifics.

On 24 August 1927, one of Maunsell's very pretty River Class tank locos was hauling an express to Dover from London, and suffered a major accident just north of Sevenoaks in Kent.  The subsequent inquiry determined that the water in the tanks started surging which caused a rocking of the loco, and the front pony truck became derailed causing the loco and the rest of its train to leave the rails.

The Chief Engineer responsible for the track was George Ellson.  Apparently he suffered from a nervous disposition and after the accident, was very suspicious of front pony trucks.

Though more likely the main factor was poor track, the 'K' class did have a reputation of instability at speed that had led to derailments, and they were quickly rebuilt as tender locos forming the 'U' class.

Maunsell was at the same time struggling with the performance of the Nelson's which he was never really satisfied with.  They had a long narrow firebox which demanded very careful firing and their steaming left a lot to be desired.

                                                  
Puzzling on what to do next, in 1933 he laid out a basic design for a Pacific.  This was based on the Lord Nelson design but had a wide firebox which should have made firing much easier.  However, the economic climate was not right.  Not only was the country still in recession but the Southern’s General Manager, Sir Herbert Walker, was busy playing with the largest electric train set in the world – the Southern Electric - and there was no money to fund new steam locos.

Maunsell also tried to get approval to build a 2-6-2 tender loco but the full-sized version was turned down for the same reason.  However that 2-6-2 design did see the light of day in model form, as 'LBSC' used it as a basis for his 3½" gauge design known as 'Betty'.

But by this time Maunsell was 65 years old and was suffering from ill-health, and the momentum he had built up started to slow; in 1936 no new steam locos were built by the Southern.  On 28th May 1937, Maunsell announced his retirement and whilst he left a legacy of some fine locos, one disappointment must have been not having a really effective heavy express passenger loco.  Sir Herbert Walker was charged with finding a successor, and being a very careful and cost-conscious manager may well have appointed one of Maunsell's lieutenants, James Clayton or Harry Holcroft.

                                                  
However, the Chairman of the Southern board, Robert Holland-Martin had other ideas, being very aware of the prestige being earned by the LMS with their Coronation Locos, and the LNER with the A4.  O.V.S. Bulleid had been Gresley's right hand man since the GNR days, and at 54 years old had a good 10+ years to serve before retirement.

Although not directly responsible for the design of a loco, he had been involved in the designs for most of the LNER carriage stock since the grouping, and also had extensive experience in the testing and development of most of the LNER express locos - in particular with the class 'P2' 2-8-2 'Cock o the North'.

                                                  
Bulleid was invited by Walker to an interview at Waterloo and to apply for the job as CME for the Southern.  As the only candidate Bulleid got the job, and took up the post at Waterloo on 20 September 1937.  For the next few weeks he had a handover from Maunsell, who finally left on 31 October.

On his appointment, and new to the Railway, Bulleid immediately reviewed the steam loco stock and became aware of the shortcomings in the steaming of the Lord Nelsons.  He ordered a redesign of the cylinders to enable an enlarging of the valve diameter, and had a multiple jet blast pipe installed.  These changes plus enlarged steam pipes made the steaming much better, and the Nelsons became good reliable work horses for the rest of their days.

Bulleid then started to think about a really powerful loco that would, once and for all, sort out the Southern's express loco problems.  He had seen the benefits of the large engine policy on the LNER but saw many express trains leaving the Southern's termini still being double-headed, so he got a team together at Eastleigh and started working on designs for something really spectacular.  This included overcoming objections from George Ellson, the Civil Engineer, still cautious from the scare when the River Class loco crashed at Sevenoaks.

Bulleid's first outline scheme was weight Drawing No. W3519 of January 1938 showed an outline of a Pacific.  But Bulleid felt that he would need more adhesion for the Dover Boat trains, so soon changed his thinking to a 2-8-2.  Diagram W3630 of June 1938 showed an axle weight of 19 tons, but George Ellson refused it on the basis that it had a leading pony truck, so back to the drawing board. 

Bulleid kept up to date on technical developments around the world, and revised the 2-8-2 design to incorporate a Hemholtz truck which provided articulation to the leading coupled axle, in line with common Continental practice.  Ellson was sceptical about this and preferred a leading bogie, and there were also concerns about the length of the loco and tender, so reluctantly Bulleid reverted to a Pacific.

The next design turned out to be the 'Merchant Navy' class loco as it was to be built; the axle loading was 21 tons which was the maximum allowed at the time.  Bulleid sought approval to start detail design work, which was given.  However the Country was by now at war, and Bulleid had a big selling job to do on the Ministry of Supply to get the materials – he achieved this by describing the locos as 'mixed traffic', although they were rarely seen pulling freight trains.  With the go-ahead given, Bulleid started laying out some basic principles of the detailed design:

                                                  
The locos must have a large steaming capacity so that never again would the Southern suffer from being underpowered.

Bulleid designed the biggest boiler he could possibly get into the loading gauge.  It had a large wide firebox with around 50 sq ft of grate area.  The boiler had an unusually high working pressure – 280 psi which added to the power available.

Bulleid decided to follow American practice and incorporate two thermic syphons in the firebox - the idea of these was to promote rapid circulation of water around the firebox.

The firebox was also welded – Bulleid was very keen on this practice.  However, it took him time to train a team of welders at Eastleigh, so the first 10 boilers were actually made at the North British loco works in Glasgow, delivered appropriately by rail.

Next, the smokebox: Bulleid had an adage 'form follows function' – this was a horrendous looking fabrication with few straight lines and lots of complicated bits in it.  (Contrastingly, the smokebox of the later 'light' pacifics was all straight lines).

                                                  
As the locomotives needed adequate power to haul all the normal boat and express trains without assistance, Bulleid decided on three cylinders as this would not only give him the necessary power, but provide a high degree of dynamic balancing.

Bulleid also wanted to ease the lives of the loco crews who had to get to numerous and often inaccessible oiling points.  He was not unhappy with the idea of outside valve gear, but did not want a conventional centre valve gear as this would be difficult to get to, or to incorporate a derived motion as used by his old boss Gresley on the LNER.

So he hit on the idea of placing all three sets of valve gear between the frames with dust protection and adequate, automatic lubrication -   in a sealed chamber, with oil sprayed as required.

His early thinking was to use eccentrics and conventional pin joints, but there were design issues so he hit on the idea of driving the valve gear either with gears or chains.  Gears were difficult to get hold of, so chains became the favoured solution.

                                                  
Bulleid also recognised that it is very hard work to drive and fire an express loco, so he wanted to ease the burden of the footplate crews here as well.  He arrived at a really good cab layout with grouped controls, steam-operated Ajax firedoors, oil boxes to feed all the axleboxes, and a steam generator to illuminate the cab controls and gauge glasses – also providing power for the front and rear lamps.

The trailing truck incorporated a three point suspension system which contributed to the smooth riding of the locos – a design used later in the BR standard locos.  The wheels were designed by Bulleid in conjunction with steelmakers Firth Brown – so known as BFB.  These reduced the weight by about 10% compared to normal wheels, and provided a more continuous support for tyres.

Access to a book retrieved from a pile of rubbish at Eastleigh revealed all the works costing data; below is a photo of the cast steel frame stretcher that supported all the valve gear, and the Works costs for machining it – it took 28¾ man hours for marking off alone, which gives you an idea of how complicated that stretcher is;  the wage rate is interesting - I calculate it to be just under 10p per hour !

                                                  
Mainframes laid out at Eastleigh, where the whole class was built – note the long sweep at back, for under the firebox.

The first 10 were built at the height of the War years, and amazingly survived the bombing runs.

Here are the moulders in the foundry with the centre cylinder mould – foreman in bowler hat supervising.

                                                  
So here, the Merchant Navy class prototype as initially built; note numbering of 21C1 following continental practice.  The whole class had cast nameplates featuring the flag and name of the particular Shipping line, but only the first two were fitted with the cast gunmetal number plates (sides and front) and large 'Southern' plates on the tender, as shown in the photo; rather elaborate considering the shortage of materials for the War.

Note the distinctive 'widow's peak' and rather ugly slot above the smokebox door to allow air to carry exhaust smoke away.  The initial curved 'Southern' casting adorning the smokebox door was likened to an unlucky horse-shoe, and was soon modified to form a closed roundel, showing the build year.  However both 21C1 & 21C2 were found to be overweight, and modifications were made before more examples of the class were built.

                                                  
This photo is of the naming ceremony for 21C1 'Channel Packet' on 3rd March 1941; with Lt Col JTC Moore Brabazon - Minister of Transport undertaking the duty, with Bulleid to the right.  After the ceremony, 'C.P.' and attending dignatories travelling in attached coaches went on a junket up the Mid-Hants line to Alresford.

'Channel Packet' and sister loco 21C2 'Union Castle' were sent to Exmouth Junction shed for testing.  A fault quickly emerged - the valve rocker on the centre cylinder broke; this was due to an imbalance of the outside admission steam chest pressures, caused by the lack of a balance pipe between each end, placing additional strain on the rockers.

The front end was soon remodelled, partly as a result of drifting smoke; the 'widows peak' was replaced with the now familiar rounded cowl at the front of the smokebox, and smoke deflectors were fitted.  The classic shape finally emerged in 1945 when the cabs were modified to give a better forward view.

But the revolutionary design was not without further issues; the design of wheels allowed oil from the axleboxes to be thrown up onto the lagging which occasionally caught fire.

And problematically 'Bibby Line' broke its crank axle – fatigue had set in around the area of the chain sprocket;  the axle broke while running through Crewkerne station causing the left hand coupling rod to come off and take out the station canopy !

Despite these set backs, soon there was demand for a lighter loco with more route availability – so Bulleid took the MN design, reduced the boiler length and cylinder size, included more fabrications and a simpler and lighter tender, and the 110 'Light Pacifics' were born.

                                                  
Nationalisation came along in 1948, and Robin Riddles and his team started evaluating locos from the 4 former companies; tests were carried out on different lines (the locomotive exchanges) and also at the Rugby testing plant.  21C22 (35022 in B.R. numbering) 'Holland America line' was put under test on the rolling road; the steaming rate got to 42,000lb per hour, about the same as the Stanier pacifics, but the maximum potential of the boiler was not reached due to oil on the rollers.  (Bulleid's boiler design was so successful that it was used as the basis for the Britannia and other B.R. standard designs.)

Moving into the 1950's saw shortages of manpower and a consequent deterioration in maintenance; the Bulleid pacifics were too complicated for the average fitter, and availability began to suffer.  This combined with heavy coal and oil use lead B.R. to question the viability of the Merchant Navy pacifics.  They concluded that rebuilding along more conventional lines would be the answer; a re-design was worked up by Roland Bond who had worked under Robin Riddles.  The valve gear was replaced by conventional Walschaerts, the cladding was removed and the result was a handsome and uncomplicated machine that performed well.

But to reflect: Powerful, sometimes temperamental, always controversial - but they so greatly lifted the image and morale of the post War Southern.  Imagine London just after the War, a grey and weary place, when into the scene comes this . . .   
'Channel Packet' heading a post-war 'Golden Arrow' express; certainly the 'Southern' now had an worthy icon for their publicity posters.