Javascript is required

A web-site by Rob Speare


34092  - 'City of Wells' in 5" gauge.

Words by Stefan Owen.

I used to enjoy 'double maths' back in the late fifties.  This put me in a classroom with a large panoramic window overlooking the playing field, with the railway track beyond.  My desk was next to the window, by the radiator, and from this position I could see the train running up and down the local branch line.  Of course I was not distracted from the maths - I loved the subject.

Being part of the Somerset and Dorset branch line, smaller engines used to do the work.  Class 2 and 3 tanks, Collets and 2Ps were the normal motive power.  I think 'City of Wells' travelled over these tracks to attend it's naming ceremony in the City, just six miles away.

I spent many years building a 2" Fowler ploughing engine.  Then I started an Ivatt tank, a native of the branch line.  But talking to a long time colleague of mine, Melvyn Marshman, changed all my plans.  We talked about this thing about the engines taking such a long time to make, versus the reducing time we have to make them in.  Why didn't I build the engine I really loved ?  Having been released from my full time employment this gave me a real chance.

So I shelved all other hobby activities, and made a start on the 'City of Wells'.  At the time I only had the Keith Wilson articles and lots of photographs, so I knew it was going to be an up hill struggle.

Having worked at Clarks Shoes using CAD CAM to produce sole mould tooling, I thought I could use the technology to help produce some of the loco parts.  As most people, I started to build the tender first.  My information of West Country engine details was very limited.  Knowing '34092' was of a later batch of engines to be made, I thought it would be issued with 5500 gallon tender.  I started to draw up the Ariel tender, but changing the design to what I knew about the smaller tender.

My first parts were soon drawn up and made into 3D components within the computer programme.  By doing this I knew all the parts fitted together giving confidence to get them cut.  I had other dealings with Pete Thomas at Polly (Model Engineering) at the time, so I sent him the DXF file.  He soon returned a kit of parts, which I could easily assemble into what looked like the tender I was after; silver soldering, riveting and bolting it together.

I've got this idea, like many engine builders I suspect, to buy or build bits and put them in store, so that one day I can have a big assembly and fit all these acquired parts together.  I have purchased injectors, safety valves and as many ancillary parts as I can afford.  My front bogie is as Ariel, except for 'chevron' shaped stretcher at the front; now finished and under a dustsheet, waiting.

The Boiler

Now what to build next ?  I thought the boiler fell into the above category, ever increasing in cost, and a storable unit.  So I made my mind up to build this, having gained confidence on two previous boilers.

As far as I can understand it's the same as Ariel only about an inch and a quarter shorter, all taken off the firebox end.  I drew the boiler up on the computer and produced some templates to make the formers.  This is when I first got a first idea of how big it was going to be.  I used dense plywood to flange the plates around, which enabled me to band saw and file to shape easily.  I used a mix of Propane and oxygen to get the copper up to temperature ready for silver soldering.

I did not use girder crown stays, but made some long copper rivets and soldered them into the top of the firebox.  The whole process took about a year on and off.

Around the time of boiler building three things started to come together.  Firstly, a friend Arthur Young gave me some link chain; I consider the chain drive to be at the heart of the original Bulleid locos.

This started me off thinking about the valve gear.  Being aware of Charlie Dockstader and his valve gear simulation programme, I spoke to Simon Bowditch and asked if he could fine-tune all the theoretical valve components.

Second thing - I spotted another West Country engine 'Lyme Regis' being displayed under construction at a Taunton exhibition, by Dave Cox.  He has focused on the valve gear, and this was the first time I had seen what goes on in between the frames of this locomotive.  Dave had worked from works drawing and achieved what I wanted to.

He spent a lot of time researching, drawing and making these parts, so I would not presume for him to give me this data on a plate.  What I did in conjunction with Dave is that I took the Charlie Dockstatter 2D measurements and CAD modelled to 3D components.  Drawings of these were sent off to Dave for him to check and correct; like student and master.

And third - I came across this web site which deals with model Bulleids – what a wealth of information to explore.  I have learned that the original light pacific is a very different engine to Ariel, or a Merchant Navy.

I found out that Rob had an interest in a similar engine to mine, and he decided to commence a build too.  This was a good opportunity to share resources, using the CAD method of manufacture.  Rob has helped by producing drawing/sketches of the stretchers, together with photos.

From these I then create the 3D CAD model, checking for fit, then create 2D files ready to send to the steel cutters, and store them in a library.  Sometimes a 3D model is required for the lost wax method of manufacture.  I ask the computer to wrap tri-angles around the model, and output a STL file (stereo lithography).  This can be sent off to have a solid physical model/pattern made to go off for casting.

This truly is bringing technology to our hobby, hopefully speeding up the whole process, while continuing my original love of maths and steam engines over 50 years later.

Radius Rod Assembly

This is another set of parts which have been a headache.  Starting with the top bracket, this is the now usual draw up, laser cut, bend and solder.  The five bearings bushes being the most difficult.  I ended up soldering them and setting the assembly onto the cross slide and drilling them all inline.  The centre expansion links were also drawn up and laser cut.  The outside parts straight forward milling, turning and filing.  The hardest bit for me was the concept of how it all fit together and works in the bracket.

The Radius Rods are another difficult part to plan manufacture and execute.  I did have some help from Dave Cox in drawing these up to what they should be made like.  Because of the size of my chain, hence valve crank journal and the front plunger position I have had to tweak the measurement to fit.  Starting with 5/8 x ½ bar I milled and filed the centre hole.  Leaving a good length of bar either end full size to enable gripping in the vice.  The rest became a series of milling, drilling and fitting (- whittling as my wife puts it).

Valve Crank Assembly

For some time I had been worried about how to make and shape this component.  I decided to 'build' the three throw crank, silver soldering the joints.  But I needed a way which minimalised any distortion.

The chains I have dominate the very little space between the frames.  So I drew the crank up using 1/8th for the webs, 7/16th long journals and 3/8th in diameter.

Drilling a row of 5/16 holes in a ½ x 1/8 inch bar 0.4 inches apart.  With a light counter sink each side and half a dozen small 'V's filed in the holes, this to allow good silver solder penetration.  I turned a shoulder on the journals so that they would be a light push fit on the journals,  finishing with a 3mm drill hole down the middle.  This allowed me to bolt all the parts together for silver soldering and cooling down after.  The result turned out very well.

The Casing

When I started this component I didn't have a clue about how to construct it.  My mind was a blank just like a sheet of virtual paper, and like the valve gear I just did not know what was going on.  What I did know was the height and width, and the outer shape.  Things I had to invent were the skeleton to support the skin, what materials to use to make it from, and how to hold it all together.

Starting with skeleton I used CAD to design the support ribs.  I took the outside shape and offset it by 10mm, and 12mm in the top corners - just a guess at these measures.  Had these U shapes laser cut in 1.5mm BMS, with the intention of wrapping a strip of 10mm wide by 1mm around the outer shape, silver soldering together forming an 'L' section.  This did not work as the silver soldering process causes too much distortion.  I tried to bend back into shape, unsuccessfully, and also lost the 90 deg of the 'L' section.

Next, I went to the local B&Q to look for some 'L' shaped extrusion.  There were several choices, 10mm by 10mm in brass, aluminium or plastic.  The brass at the time was about £11.50 per metre and the aluminium £2.75.  So I decided to buy a couple lengths of the ali to try.  I set about cutting 'V's in one side to enable bending around the corners of the former, then securing with some 8BA screws.  This method produced some good strong and fairly accurate ribs.

Weighing up what to use for the skin was difficult.  Thin copper I know I could bend round O.K.  Some of the questions I could not answer were e.g. How thin/thick?  How would expansion affect the fit of this outer skin ?  How would copper take paint ?  Also possibly expensive.  Another requirement of the design is that the casing should be able to be removed easily.  Similar questions I ask myself about thin brass, a more conventional material for this job. 

But a colleague of mine had just been working with tin plate, and let me have an offcut to play with.  This piece was .5mm thick and quite rigid, would take paint well and have minimum expansion, this seemed favourite in my mind so I ordered a sheet.  Just did not know what to hold it all together with, as it would be impossible to use scaled bolts as full size practice.  The smallest manageable screws I could find was 14BA, this meant drilling lots of 1mm holes.  14BA is still over scale but I have spaced them out more, and think I can just get away with this.  Taking the bull by the horns I ordered 500 screws and nuts.  With my CAD package I drew up a panel with all holes marked, then sent off to be laser cut, leaving all the holes to drill in the aluminium ribs.

Using centre lines and marks I draped and wrapped the tin plate skin over two formers, holding with the screws as I went.  This produced a pleasing strong shape giving me confidence to carry on and make the rest of the panels.  To secure the casing to the chassis I am using 6BA thumb screws through the brake hanger brackets.