Saturday, October 23, 2021

How I constructed a Kerr Stuart 0-4-2T loco using 3D printed parts

Unlike most of my previous loco builds, this one started with the motor block. A fellow modeller had kindly donated the newer version of a Bachmann Lyn motor block while I was struggling to get an older version of the motor block to work reliably (see How I converted a Bachmann Lyn into a Schull & Skibbereen loco). The new block needed some remedial work to get it up and running but once that had been done (thanks to some spare parts from Bachmann in the USA - the postage was almost as much as the cost of the parts), I then had a rather fine and, dare I say it, well engineered motor block in search of a loco.

By the way, in my opinion, the new version of the motor block is far superior to the old one. Firstly, the two oversized worms have been replaced by a single smaller worm and a two-stage reduction gearbox; the quality of plastic used for the motor block is far higher (it doesn't crack and disintegrate); the motion is better engineered and the fly cranks are secured to the wheels much more reliably.

So, I searched the web for a suitable narrow gauge locomotive with outside frames which would suit the dimensions of the motor block. I couldn't find a three-foot gauge loco which fitted that description but I did track down some images of Kerr Stuart Skylark locos which to my mind look very attractive - even though there were two-foot gauge.


Clearly, the loco drivers on the Lyn motor block are far too large and I wasn't sure how it would look regauged to three-foot. I then came across a photo of one of the Kerr Stuart locos which ran on the Metropolitan Water Board Railway which looked like a far better proposition.


Although the driving wheels on the motor block were still too large, I felt the chunkier loco body could easily accommodate them. There was also a line drawing of the loco on the same web page, which clinched it for me.

Having decided on the prototype, I started where I almost always start with my scratch-bash-builds - with the running plate. 

There was no front, rear or plan elevation of the loco and so I based its width on similar sized locos I have constructed previously - ie 100mm. The running plate was drawn in TinkerCAD, with a cut-out in the middle to enable it to sit snugly over the motor block.


 NOTE: TinkerCAD is a free online tool which enables 3D images to be drawn using simple building-block style tools which can be combined together. As it has been designed for children to use, it makes an ideal easily accessible introduction to the delights of 3D drawing. Despite its low entry learning curve, it has the potential to draw quite complex and sophisticated shapes. For more information on how to use TinkerCAD see my tutorial.

The back of the motor block was trimmed down to enable the running plate to sit evenly - in line with the bracket supporting the motion .......

.... and the running plate was printed and test-fitted.

Pleased with how this looked, I decided to work on the smokebox. Normally, I start with the cab and work my way forward, but I was interested in how the front end would look.

The smokebox was drawn first, so I could use its dimensions to guide those of its support.....

.... which was drawn next.

NOTE: The main reason for drawing the smokebox first was so I could use a copy of it, turned into a 'hole' shape, to cut out the curve on the top of the support.

These two pieces were then printed out and positioned on the running plate - not yet fixed in place until the boiler and firebox had been printed.

Next, the boiler was drawn. At first, a small cut-out was made for the wiring, but I quickly realised that I would need more space inside the loco and so ......

...... the boiler was halved at the point where it would be hidden behind the tanks.

 NOTE: I later hacked a bit more off the side of the boiler (with a razor saw) when the tanks were printed to provide even more room for the 18650 cells to be squeezed into the tanks. I have now adjusted the drawing of the boiler to take account of this later modification if you want to download and print this loco for yourself.

Before fitting the boiler to the smokebox, I drew the firebox. The basic shape was drawn first .......

.... and then details were added to the backhead. Having no drawings or photos of the interior of the cab, I improvised.

At first, I tried printing this out as a single unit .... but my optimistic assumption that the printer would cope with creating the backhead without support was misplaced. The spaghetti which resulted caught on the printhead and dragged it off the bed.

So, the firebox was redrawn as two separate pieces .....

.... which were printed-out and test-fitted to the running plate .........

..... before being glued into place with the boiler and smokebox assembly.

I was now ready for the cab. The four sides were drawn separately. The sides were easy enough to draw - a flattened box shape with two large circular 'holes' being used to create the curve of the doorway. This was done by eye, rather than trying to compute the radii of the two circles. Once one side had been drawn ....

.... it was duplicated and flipped to form the opposite side. This is one of the great advantages of 3D drawing and printing to my mind; the ability to duplicate parts on screen.

The back of the cab was drawn next.

This was then duplicated and the firebox shape copied and turned into a 'hole' to create a correctly-sized cut-out to enable it to fit snugly over the firebox. In reality, a small amount of filing was needed to allow it to fit perfectly as the printing process tends to increase the size of parts by around 0.1mm.

The sides of the cab were printed out, .......

....... assembled in situ on the running plate and glued into place.

Next came the side tanks. These were drawn as two complete units. Previously, with my other 3D printed loco (see How I constructed a Schull & Skibbereen loco), I have drawn the tanks as a set of separate parts and glued them together, but I figured out these tanks could be printed upside down without the need for additional supports. The tank was duplicated and flipped for the opposing tank.

The decision to print them upside down proved successful .......

..... and a couple of filler pieces were also drawn and printed to block off the space beneath the tanks.

The tanks were then glued on to the running plate beside the boiler.

Now came the interesting part - and the most challenging in terms of drawing in TinkerCAD - the chimney, dome and safety valve.

Fortunately, having already drawn these parts for the Schull & Skibbereen loco, I was able to duplicate the chimney and dome, and resize them for this loco.

Getting the flare at the base of the chimney and more particularly the dome is a bit of a challenge in TinkerCAD. I've not entirely cracked it but the judicious application of some filler can help to disguise the join.

The dome was divided in half so that it would print more easily.

I was using PETG filament for this build as an experiment. I found it is much harder to get a smooth finish when filing and sanding than PLA.

Similarly, PETG doesn't seem to handle fine detailing as well as PLA. The rim around the top of the chimney came out slightly ragged.

The safety valve needed a fair amount of additional work. The tubes are taller than those on the Schull and Skibbereen loco and the whistle is mounted on another vertical tube beside it. I decided to use a piece of copper tubing for the whistle mount as I wasn't confident the PETG filament would be able to cope with it.

Again, you can see that PETG filament doesn't produce a particularly clean print, ......

..... and so some filler was needed to tidy up the imperfections.

The drawing and the photos didn't really show the tank fillers. They appeared to be rounded with a rim but I couldn't see if or how they were hinged, so once more I used my imagination. I drew them in two parts which could be slotted together ......

..... and was pleasantly surprised with how well the were printed.

Once assembled and glued into place they, to my mind, looked quite appropriate.

Next, I turned my attention to the trailing bogie. Fortunately, there is a suitable pivot for it provided on the Lyn motor block. This was sliced off (with my ubiquitous razor saw) and repositioned slightly further forward on its mounting to provide more swing room for the bogie.

The bogie was then drawn and printed out, ..........

..... before being mounted on the pivot with a collared self tapping screw (rescued from one of the Lyn bogie pivots) and a pair of 20mm diameter steel wheels on a 3mm axle (from IP Engineering) threaded through the axle mounts. 

 The weight of the steel wheels seems to keep them on the track but I might need to add some lead or a spring to the bogie if this proves problematic later.

New side frames were then drawn. 

It took me a while to figure-out how and where these could be attached. Originally, I wanted to glue them to the underside of the running plate, but this would have prevented them from providing a backing for the cylinders. I could have attached them to the to motor block, but I had removed the rear mounting pillars when the block was modified to take the running plate. In the end, I decided to attach the frames to the underside of the bracket for the rear bogie ......

..... and the underside of the support for the cylinders.



 Rather than drawing the roof in TinkerCAD, I decided it would be easier to make this from plasticard. A sheet of 1.5mm thick plasticard was cut to size and a couple of 2mm thick curved supports created to fit inside the cab.

At this stage, the loco was wired up to a Deltang Rx65c and three li-ion 185650 cells and given a test run to make sure the frames didn't rub against the fly cranks

Although she ran well, the video threw up some issues with the appearance of the loco which I felt I needed to resolve. The boiler seemed to be slanted upwards towards the front of the loco and she seemed to be sitting far too high above the track.

 The raked boiler was fairly easy to resolve. It was removed and a couple of millimetres sliced off the bottom of the smokebox mount (the drawing of the mount has now been adapted to overcome this issue).

Lowering the body was a much more difficult problem. In the end, I hacked more off the motor block - ie lowered the mount for the running plate at the back by 7mm .......

.... Lowered the mount at the front by the same amount .......


 ....... and removed 7mm from the top of the valve chests. I also sloped them down towards the front to better reflect the shape of the valve chests on the real loco.

I also carved some slots in the running plate to allow it to slot over the motion bracket and to prevent the wheel flanges from rubbing against the running plate (highlighted in red). 

I have provided two versions of the running plate and the side frames so you can choose whether to have the first or the modified version.

I drew sandboxes for the running plate at the front of the loco .....

..... but when I printed them out with PETG, they were very disappointing (on the right below). Fortunately, my reel of PETG was almost exhausted and so replaced it with PLA+ which produced a far crisper print (on the left).

These were mounted on the running plate.

 The smokebox door handle was also drawn ......

..... and then printed out. I doubt PETG would have been able to handle the delicacy of this part but PLA+ seemed to cope well with it - being a brand newly opened pack of filament might also have helped!

Rather than drilling holes in the bogie wheels - I doubted I had the skills needed to ensure they were equally spaced - I decided to print cosmetic inserts thereby ensuring the geometry was perfect.

These were printed out and glued into the recesses in the wheels.

The photos of the locos showed their buffer beams had been extended and so pieces for these were drawn with recesses to allow for the fitment of my bespoke couplings

It later transpired that, because I had lowered the body, the couplings needed to be mounted higher up and so the extension pieces were adapted and the buffer beams slotted further up to take the couplings.

Finally, covers for the valve chests were drawn, .......

...... printed and fixed to the tops of the valve chests. Although these are barely visible, I felt the loco would be incomplete without them.

The loco body was then given a couple of coats of Halfords grey primer from a rattle can aerosol .....

This was lightly rubbed down with fine grade wet and dry emery papers. Filler was applied where necessary (mostly around the base of boiler fittings and where the printing process had created some indentations). Another coat of primer was added followed by a couple of coats of Rover Brooklands Green (my loco livery) from a Halfords rattle can.

The smokebox, running plate and frames were given a brushed on coat of black acrylic, the interior of the cab was painted cream with acrylics and various fittings were picked out in copper or gold acrylic paints. The buffer beams were painted scarlet.

A 3D printed driver was added and the backhead details picked out in various appropriate colours

Brass handrails were added to the sides of the boiler (using whitemetal handrail knobs from Garden Railway Specialists) , ......

...... the front of the smokebox ........

.... and the cab doorway

Strips of lead flashing .......

..... were glued between the frames at the rear of the running plate ........

 ...... above the cylinder support bracket .........


....... and squeezed into spare space in the tanks

A small encapsulated speaker (from Rapid Electronics) was glued to the roof of the cab and connected to a Premium MyLocoSound soundcard which was wired up to the chuff cam in the Lyn motor block.

The loco was then coupled up to a short train of goods rolling stock and given a test run.

So far, I am pleased with the performance of the loco, though I still need to tweak the settings on the sound card to improve its response when pulling away and slowing to a halt.

The motor block is a great improvement on the older version. It seems to be smoother and a lot more powerful - which I have attributed to the replacement of the two large diameter worms with a smaller single worm.

I haven't had a chance to run the loco extensively to check the performance of the trailing bogie over a range of situations. Up to now, it seems to cope well with the undulations of my track though I have had one derailment of the bogie when the loco was running backwards. I'm not yet sure if this was just because of debris on the track or whether the wheels on the bogie are struggling to keep contact with the rail.

There are still a few details I want to add to the loco - the photos of the real thing show that re-railing jacks were a common feature on the running plate, the whistle mount needs more detail and I'm trying to decide if adding Walschaerts valve gear is going to be worth the trouble it would take to construct and install it. My loco is not an exact replica of the original locos so I could argue that modellers' licence applies to my interpretation of these fine looking locos.

I have decided to name the loco 'Bunbury'. Although this name has already been assigned to another loco, it would be fitting to pass the name on to this loco as one of the originals was called 'Sunbury'.

The files to make this locomotive are available on the Garden Railway Forum and Thingiverse 3D printing file sharing website. NOTE: You need to subscribe to either of these websites to be able to download the files.