A West Country class Light Pacific in 2.5 inch gauge.
An interesting new build in a smaller gauge - contributed by 2½" Gauge Society member Doug Rundle
Despite being brought up in GWR territory, I have always had a soft spot for the Bulleid West Country Pacifics in their original form.
About thirteen years ago my thoughts turned to this project and one of my original views was that I did not want to build bigger than 3½".
Coming across the National 2½" Gauge Society, I picked up the list of designs that Chris Barron offers, and there amongst them was a West Country Bulleid Pacific.
However, for one reason or another I decided that instead of building to that drawing I would design from scratch.
My intention is to produce a model that, when viewed from a distance of a few feet, is a fair representation of a WC, with all of the most prominent features reasonably accurately represented.
Discovering that WCs varied in design so much during their life (one of the factors, strangely was where they were built, Brighton or elsewhere for instance), I am basing mine on the last few built, 34100 onwards.
I have yet to decide on the number, but it is to be called "Marytavy".
The first picture is the clue.
My Grandfather was organist for the Devon Church on the tor, Brentor St Michaels; and he lived between Brentor Southern railway station and the Marytavy Great Western station.
As West Country pacifics 'Lyford', 'Brentor' and 'Tavistock' existed, and because my engine is not based on one particular prototype, my engine is named after the village that has a lot of very pleasant memories for me.
In the picture, a bit further on in the direction the loco is travelling, the Southern track passes the back door of Marytavy station.
I use several sources for information, the main one being a series of articles that were published in the now defunct "Model Railways" magazine.
I came across it by chance but it is "Portrait of a Locomotive - 13. OVS Bulleid's West Country Class, Southern Railway by Bernard Wright, Drawings by Russell S Carter."
This series started in July 1974.
The detail is superb and the line drawings with principle dimensions are as good as it gets.
Also, "The Southern Pacifics. Bulleid's Radical Design" by Roger J. Mannion; "The book of the West Country and Battle of Britain Pacifics" by Richard Derry; dimensioned drawings by F J Roche, and various other printed references.
As I live in Hampshire I am not too far away from various full size originals that are or were in various stages of overhaul or restoration.
Additionally a large quantity of Bulleid drawings are now safely stored in the National Railway Museum from which copies can be obtained, although I use this resource as a last resort.
Where would we be without the computer ?
For the design and drawing work I started out using a 2D CAD programme, but as the project became more complicated 2D was becoming more than a bit tedious, so I purchased and am using TurboCad 3D CAD software.
One thing that perhaps I should mention is that I want to incorporate chain driven valve gear working on all three cylinders.
In the other gauges builders have successfully achieved working systems, and the larger one goes the easier it is to close in on full size.
At 2½" gauge I accept that it is not going to be a true representation, not even getting close; but I felt that as I am building one of the classic locomotives then I wanted at least a strong flavour of one of its major features.
I am sure that when running there are going to be limitations, so I am not going down this path with the expectation of having a flawless valve gear; there are going to be issues and problems.
One major point of compromise is that in full size the drive for the piston valves is taken through a rocker system, but in this size it just is not practicable, so I am scheming a direct drive to the slide valves.
The chain that I have sourced is of 0.1475" pitch.
Progress on Drawings
Various 2D design layouts and sketches, including the Bogie, Trailing Truck, Driving wheels, Manifold, Regulator, Slide Bars, Frames, and some individual components.
In 3D design, the main locomotive assembly comprising most of those just listed, and the motion and valve gear design.
It is now easier to just create a single 3D model, adding to it as the project progresses.
At the moment the 3D model comprises 47 layers and the list is growing.
The boiler is part of this and has been individually detailed, and then subsequently approved by my Society boiler inspector, and an authority from the National 2½" gauge Association.
Regarding the main frames, on this gauge running with curves of smallish radius, I am concerned that the trailing truck wheels will foul the main frame.
Unless I really have to, I don't want to put cut-outs in the frames as they are not present in full size.
The plan is to cut the frames between the rear driving axle and the rearward periphery of the rear driving wheels, then to attach to the inside of the frames what amounts to frame extensions.
I propose to make the frames from 2mm thick steel – probably Ground flat stock - this means that the frames underneath the footplate will be 0.158 narrower than further forward.
From a small distance it should be difficult to tell that there is this change to the frames.
Similarly, at the front the frames are visible beneath the smokebox door quite prominently, and I want to achieve roughly scale width and thickness.
This can be achieved by cutting the main frames in the region of the outside cylinders, and adding 1.5mm thick frame extensions attached to the inside of the main frames.
Work in progress: Motion and valve gear design.
After this, the smoke box where I have already established as part of the boiler design, that I can fit in a Doug Hewson style superheater.
I want to retain one of the classic features of Bulleid's Pacifics, the large diameter chimney, so will be looking at Lempor draughting to achieve this.
Components Manufactured
Front Bogie :
This is one of the features that stand out – the side control spring housings and the front stretcher in particular.
Using various sources I have ended up with what is really a half size version of Keith Wilson's "Ariel" that has been WC'ized.
Apart from interconnection detail to the main frame this is complete.
Trailing Truck :
This is a wonderfully curvaceous assembly that was very interesting to build.
On full size the truck frame is a welded channel construction with not many straight lines.
The components have been cut out of plate so that it will look all right, not that there is much to see anyway.
The main curved frames are water jet profiled.
Wheels :
Machined from castings supplied by the National 2½" Gauge Society.
Boiler Fittings, which include :
The Manifold,
Regulator assembly - a half size "Sweet Pea",
Safety Valves - to Gordon Smith internal design,
Firehole doors, and
Tender water valves (needed one for another loco so decided to make a batch).
An update - 22nd Jan 2020.
Looking at the web site the other day, I couldn't believe that I had last contributed nearly four years ago.
At the time I was starting to think about the valve gear design, but managed to make some progress with the slide bars and brackets, and am now am making progress with the various boiler components that will be soldered to the boiler.
The slide bars were straight forward, but the brackets were more interesting.
As they are a prominent feature full size I wanted then to be visually correct.
Using the NRM drawing (E46340 of 1952) I redrew them, then had a pair cast using the 3D printed pattern lost wax method using Cro Fittings.
Adam really did me proud and I intend using this method for other components.
There are some faces to be clean-up machined but that should be all.
Meanwhile the valve gear started to take over and after three or so years of frustration, trials and tribulations and one or two eureka moments, I think I have a valve gear design that will work, perhaps.
The design aim is to try and imitate the spirit of Bulleid's version of Walschaerts valve gear as applied to the original design of the West Country Pacifics.
However, being 2½in. gauge brings restrictions to the design, particularly in material strength, and especially with regard to the swinging link in the cylinder.
Prototypically the system is outside admission but this design is the opposite.
Inside admission in the model removes the need for the swinging link and the valve rod gland is at exhaust pressure so reducing the friction loading.
After the usual initial thoughts and looking at what other people were doing in this gauge, I decided that I would incorporate Henry Greenly's double port piston valve.
This design has been successfully used by an innovative builder in this gauge, and brings the advantage that the valve bobbins are longer than usual so making for improved sealing.
Basic dimensions are derived from a method described by WD Hall using spread sheet calculations.
CAD in 2D mode was then used to create a theoretical mechanical layout eventually ending as thirty six layouts for different angular positions of the crank.
From this data a four times full size cardboard, drawing pin and plywood model was made just of the valve system, one of the three sets that is.
This was fine to prove the basic design but, as to be expected, just was not accurate enough and also I could only relate valve movement to crank angle but, as I later found out, it is far better to relate valve movement to piston position as a percentage of piston stroke.
Anyway from the information gleaned from this model a further set of CAD drawings was created to refine the design details.
I then decided that something much better than cardboard, drawing pin and plywood was needed, so a twice full size "proof of concept model" (POCM) was made from flat aluminium extrusion mounted on a 6mm thick aluminium plate.
I was able to mount this on the milling machine table and accurately locate the various critical holes.
This time a representation of the crank, connecting rod and cross head was built in with the crank connected to the valve crank by means of a small toothed belt.
The full movement of the cross head was calibrated with marks at accurate intervals of 10% stroke.
On the valve side the eccentric rod and anchor link were adjustable so that various dimensional combinations could be tried to refine the design and as percentage piston travel the link swing in 5° intervals was marked on.
The link was a simple disc with holes for a removable pin at the critical link features, e.g. full gear, 25% cut-off and mid gear.
Once this had been optimised a further thirteen layouts were CAD drawn, to accurately verify the valve events using information taken from the model of valve positions relative to percentage of piston stroke.
It was really gratifying to see how close the results were to the results from the model.
My CAD programme does not have an animation function which may have made the process a little easier, and it would appear that the simulations are available for Walschaert's valve gear but do not model my special version.
The final design is certainly a compromise that gets reasonably close to the requirements of a valve gear as specified by D.L. Ashton and others, and with more work it could be refined even further I suspect, but I think and hope that I have gone far enough.
One aspect of the design that I have not worried about is reverse gear, simply because I think it extremely unlikely that I will ever want to run my little engine in reverse.
With regard to the valve gear design special mention should be given with thanks to Simon Bowditch for patiently acting as my mentor in the later stages of the design process.
Now that the valve design is settled I am concentrating on getting the boiler built as a kit of parts, that a professional boiler maker will stick together for me.
These components were supposed to have been completed by last winter but, hey ho !