Confessions of a 3D bodgeller

Regular readers of Garden Rail, might remember I have previously shared my experiences of scratch bash bodgelling rolling stock for the Peckforton Light Railway. My approach involves keeping costs down by utilising cheap motor blocks or wagon chassis, and either converting off-the-shelf models or constructing bodies from Plastikard and 'found' materials such as bottle tops and Biro tubes. I have even turned my hand to mass-producing parts using resin castings for a rake of identical open wagons.

From time to time, fellow modellers have generously volunteered to produce bespoke parts for me using their 3D printers such as, a set of layshaft pulley wheels for driving various pieces of machinery in Peckforton sawmill.

I constructed an ex-Clogher Valley railcar and needed to equip it with seats. Rather than dusting off my resin casting skills, I wondered if one of my 3D printer-equipped mates would produce something. The difficulty was that I couldn't find any 3D drawings even vaguely resembling the seats, and felt it would be unreasonable to expect a fellow modeller to draw as well as print them. Time to have a go at drawing them myself.

I downloaded a free version of SketchUp and, as is my wont, muddled my way through producing a couple of drawings which sort-of represented the seats and seat backs. At about the same time, I happened to stumble across an eBay listing for a new 3D printer in kit form for just under £70. A couple of days a large box arrived. The blurb insisted it could be assembled in half an hour. It took me two days to figure out what went where, confounded by the fact that the assembly instructions were for a different model of printer. To my amazement, my drawings of the seats actually printed and, within a few days, I had produced sufficient seating to complete the model.

But now what?

I downloaded and printed out a few small items from Thingiverse.com, the free online repository of 3D drawings, but if the printer was going to serve its purpose for future projects, I needed to hone my skills at draughting bespoke objects. Don’t forget, I am at heart a bodgeller – and precision drawing is not one of my strengths.

My next major project was the construction of a Boneworks building to sit alongside the track at Bulkeley Station. To my delight, during the interwar period there was a Boneworks producing glue, gelatine and fertiliser located within sight of Beeston Castle. I shifted its location to Bulkeley so it would be adjacent to the railway and discovered a much more attractive Bonemill building located in Stoke on Trent, rather than the ramshackle collection of corrugated iron structures housing the Boneworks in my area. I reduced the dimensions to fit the space available, but tried to retain the general character of the building. The model was constructed from, a scribed foamboard shell with Plastikard slates. I looked forward to using the printer to mass produce the somewhat unique window frames and the various styles of ridge tiles needed for the hipped roofs.

At first, I drew the frames with SketchUp using my newly acquired skills. However, despite the drawings being fairly basic, I just couldn't get them to export successfully as .STL files which the slicing software for the printer would recognise. I kept being informed that some of the shapes in the drawings weren't 'manifold' which in plainspeak meant they weren't fully formed – 3D printers just aren’t adept at printing things which aren’t there. I spent a couple of days trying various ways to fix the drawings, but without success. A fellow modeller on one of the online forums offered to fix them for me and then talked me through what he did. The explanations went in one ear, rattled around a bit and then floated out of the other. I just couldn't understand how and why something which looked absolutely perfect on screen turned out not to be!

And then I discovered TinkerCAD.

TinkerCAD is an online 3D drawing environment which is unlike any other. Whilst most 3D drawing packages are based on conventional 2D computer aided design (CAD) software using lines to create 2D shapes which are then extruded into the third dimension, TinkerCAD uses 3D shapes from the start. Shapes such as cuboids, spheres and prisms are placed on the workplane and then manipulated, joined or sliced. The fact it has been specifically designed to produce drawings for 3D printing helps ensure it is more or less bullet proof. It was also developed for children to use, which might have a bearing on why it appeals to my mindset as a former primary school teacher! By comparison with SketchUp, I was able create a printable drawing of a window and a ridge tile in under an hour - to me, it just made sense! I suppose drawing with simple shapes on screen chimes with my approach to model making. As a bodgeller, I am quite happy putting things together and then tweaking them if they don't quite fit. Before long, I had a collection of various sizes and styles of window frame and ridge tiles to complete the Boneworks.

So, what next?

Having gained confidence with creating fairly straightforward drawings for the printer, I sought out further projects which would be suitable. I decided that the additional running-in nameboards for a couple of my stations were ideal candidates, followed by station fencing and a set of hinged level crossing gates, complete with lanterns.

So far, so good. However, these items were little more than enlarged, distorted versions of window frames. Time to challenge myself further with a couple more open wagons. Although I still had the moulds, I speculated that I now had sufficient knowledge to try 3D drawing and printing the sides and ends to sit atop the cheap but effective Hartland Loco Works (HLW) mini-series wagon chassis.

The components proved to be embarrassingly easy to draw in TinkerCAD. After all, if you think about it, the sides and ends of an open wagon are little more than a series of rectangular planks held together with a set of rectangular iron straps.

I then discovered that HLW had shut up shop, so my source of cheap chassis had dried up. There was nothing else for it but to have a go at designing, drawing and printing my own version. Whilst I was at it, I decided to make changes to the hangers, axle boxes and springs to make them more appropriate for UK based narrow gauge railways. Buoyed with enthusiasm, I also added a simple representation of brake gear - something sadly missing from the HLW chassis.

Drawing the chassis was a bit more challenging, but I found my bodgelling experience was incredibly useful. Producing drawings in TinkerCAD is very much like putting together a model from individual bits. A W-iron is basically a couple of triangular prisms with a bite taken out of them. A leaf spring is a series of curved plates nestling together. An axlebox is a squashed cube with a couple of bits and pieces attached. I find it quite therapeutic modelling with 3D objects on screen – and a lot less messy. I can even do it while watching TV with my other half – without the need to Hoover up the bits or worry about spilling glue! A really clever feature of TinkerCAD is the ability to turn any object into a “hole” which can then be combined with another shape to hollow it out or remove a specifically-shaped chunk from: very handy for creating an axle-hole or taking a triangular bite out of a W-iron. Solving each new problem of creating the next shape when drawing a complex object is part of the joy.

At this point, I realised I was no longer restricted to adapting prototypical items of rolling stock to fit whatever off-the shelf-chassis I could acquire. I now had the wherewithal to design, draw and print an entire model from scratch. As a long-term fan of the Southwold Railway frustrated by the paucity of models and kits available in 16mm, or better still 15mm scale, I didn't hesitate. A four-plank Southwold open wagon was followed by a short wheelbase van, a three-plank open, a two-plank open and a long wheelbase van. The beauty of 3D on-screen modelling is that components from one model can be copied and pasted into another. Why waste effort drawing identical leaf springs over and over again?

Inevitably, I had to tackle drawing, printing and constructing a locomotive. To this end, I was fortuitously offered a Bachmann L&B 'Lyn' at a very reasonable price, and quickly converted it to battery power and radio control, but was a bit disappointed with its performance. The outswing as it rounded my tight curves meant I wouldn't be able to install my usual modified hook and loop couplers and, furthermore, it just didn't look right. Creating a model of a 1' 11½" gauge loco which could run on 45mm track has clearly led to compromises, so, I sought out a proper 3’ gauge locomotive which could make use of the Lyn’s motor block and outside frames. I found a Thomas Green 0-4-4T loco which ran on the Schull and Skibbereen Railway which looked promising, though I preferred the body styling of a Nasmyth & Wilson 4-4-0T which also ran on the same railway. I decided to marry the 0-4-4T chassis to the 4-4-0T body.

The drawing inevitably took longer, but it's surprising how a few basic shapes can be joined together to produce a reasonable representation of a steam locomotive. The most challenging aspect was producing the flare between the base of the dome and the top of the boiler, but I am particularly pleased with how the safety valve turned out. Here my scratch bash bodgelling experience enabled me to build on screen a series of shapes which ended up becoming something which I think actually looks like the original - though I shortened it slightly to better fit the chassis!

I am pleased with the outcome, but the loco struggled to pull even a modest train. The motor block had been modified by a previous owner to take a larger electric motor which resulted in the can of the motor rubbing against one of the gears. I bodged a solution to the problem, but realised a bespoke clamshell case to hold the motor more securely was needed. Once more, TinkerCAD came to the rescue. One of the great virtues of 3D printing is its ability to produce quite complex sculpted shapes - such as the two halves of a motor block clamshell.

So, where do I go from here?

To my mind, my 3D printer and the associated skills I have now acquired to design, draw and print my own models opens up all sorts of possibilities. As Del Trotter would say, "The World is my Lobster".

Over the years, there have been numerous innovations in railway modelling materials, tools and techniques which have been regarded with suspicion by some and welcomed with open arms by others. I can still remember when we had to use Cow Gum and mysterious pots of Cascamite bubbling on the woodwork classroom stove, now, we have glues such as PVA and cyanoacrylate superglues which have revolutionised how materials can be joined. Then came Plastikard, and more recently PVC foam board making inroads in the production of buildings. And how about the dark arts of brass etching, laser-cutting and CNC machining? These techniques have now become accepted procedures for the production of components and parts.

A 3D printer can be purchased for a relatively modest sum and the skills required to operate them at a basic level are fairly easy to acquire. Even a ham-fisted bodgeller like me can use on-screen 3D modelling tools such as TinkerCAD to design and produce their own models and, more excitingly, share them with others. My HLW wagon chassis is on Thingiverse.com and the files needed to recreate my Southwold goods stock, on the GardenRails.org forum for anyone to freely download and print out. In the same place is an expanding library of files and links to items appropriate for garden railway scales and modelling provided by other modellers which I hope you will use and to which you will contribute. This ability to create, tweak and share files is yet another impressive and rewarding feature of 3D printing technology. Unlike Del Boy, sharing resources in this way will not make us millionaires, but I certainly believe this technology and the mindset it engenders has the potential to bring the garden railway community even closer together.