| My professional interests |
|
Computing Law Education Science and research |
| My leisure interests |
|
Martial arts Heritage railways Garden railways Motorcycles DIY |
| Downloads |
|
Linux downloads Windows downloads Java downloads Perl downloads Home automation downloads |
| About me |
|
Home & family My CV |
| Site info |
|
Contact the author Download policy Keyword index |
Note: Since I recently moved house, work on this railway, and these pages, has ceased. I am currently working on a new G-scale railway, taking into account what I learned from this one.
Choice of buildings and structures
Commercial garden railway buildings are expensive. I mean really, really expensive; falling over expensive. What's more, those available in the UK tend to be either German or American in style. This is fine if that's what you're looking for, but I prefer a more neutral appearance. The commercial products use two basic construction materials: polystyrene or similar (expensive) or wood (very expensive). They may be available in kit form, which takes about 10 percent off the price, and unpainted, which takes off another 10 percent. However, you're still looking at about 200 quid for a rudimentary railway station, and you'll still have to paint and assemble it yourself. But apart from the price, the real problem with these commercial buildings is that they don't look very nice (in my opinion). One notable exception (other suggestions welcomed, of course) is the range from injection-moulded plastics firm Censa Ltd. Their buildings are British in outline, and look extremely impressive when properly painted. They are available from Railside Models, who can also supply pre-painted versions -- but it's `price on application', so be prepared to mortgage something.Whether you buy or build your own, a fundamental decision is whether the buildings will be a permanent fixture, or whether they'll be taken indoors in extremes of weather. In the UK we don't have the climatic swings of other countries, but the effect of frost one day, rain the next, and scorching sun the next should not be underestimated. Plastic buildings will probably stand up to the elements better than wooden ones, but unless they are coloured in the mould (rather than painted), there is still a risk that paint will crack or run. There are plenty of real (that is, full sized) wooden buildings in the UK, which suggests that it is not impossible to weatherproof wood.
Scratch-building in wood
I chose to build all the structures in our garden railway from scratch in wood, to my own design. Perhaps `design' is too dignified a word for the process involved (mostly trial and error). This is quick and, if done my way, cheap. For example the simple trackside depot shown in the photos below took a total of six hours to construct, including painting and weatherproofing, and cost about £15 for materials. Despite its frame being made of balsa wood, it is strong enough to withstand being dropped and leaned on. By way of tools you will need a saw, a mitre block, some glue, sandpaper and, er..., that's it. Of course, I am not interested in perfect realism, but instead a quick and cheap way to make tolerably decent buildings. Proper modellers should clench their teeth at this point, if there are any still reading.
The construction technique I use is based on two basic observations.
- Balsa wood is very easy to work - it can be cut with scissors, for example - but has no significant structural strength. Pine, on the other hand, is quite strong, but requires the use of proper tools.
- PVA glue bonds very well to most woods, but takes ages to set hard. Superglue, on the other hand, although it does not bond very firmly, sets in seconds.
Once the basic balsa frame is in place, I then `clad' the building with pine stripwood of 3/8'' x 1/8'' cross section. To scale this would be about 8'' x 2.5'', which would be rather large for non-structural wood. However, it doesn't look too bad in practice, and is structurally very stong when all the pieces are bonded. You need to make sure that the stripwood is bonded to other stripwood, and not just to the balsa chassis, otherwise it will fall to pieces if knocked or dropped.
My approach to bonding the wood is to use a mixture of Superglue and PVA adhesive. Each piece of cladding wood gets a couple of dabs of superglue and a thin film of PVA to bond it to its neighbours. The Superglue sets in about 5 seconds, while the PVA takes about an hour. With this combined approach, the bonds have set well enough to work with in a few seconds, and will set rock hard in due course. Note that PVA glue is available in both waterproof and non-waterproof varieties; if you use the non-waterproof variety outdoors, it will eventually dissolve.
For window and door appertures, I use 1/2'' balsa strip, as these don't need to carry any stresses. I generally have the strip run all the way from the bottom of the frame to the top, as this allows the cladding to be supported on the window or door apertures as well as the frame. You can clad all round and then cut the appertures, but this seems too lazy, even by my poor standards. The door and window frame trimmings can again be made in balsa; 1/8''x1/8'' looks about right, although in fact somewhat overscale. Guttering and pipework is made from 3/8'' pine dowel - available from timber yards - and stained black. For a real metallic finish, you could use black enamel, which does bond quite well to clean wood. However, enamel takes about 12 hours to dry, so it needs to be the last thing done.
So the basic structure ends up looking like this:
Although it is made entirely from stripwood and balsa, the finished structure is strong enough to stand on if it's properly bonded. If you're really lazy, you can form the joints so that they butt together (as in the example above); not only is this quicker to assemble, it means that you need a larger number of pieces all the same size, so cutting is quicker. However, it is more realistic, and stronger, to lap the joints, like this:
Many of the trimmings and accessories for buildings can be found laying around the house, especially if you're a DIY enthusiast. The copper conductor in mains cable, for example, makes passable brass pipework. However, on reflection I'm not entirely sure that a baked-bean can makes a realistic water tower, despite being the right size and shape:
Weathering and ageing
I have to confess to being in two minds about the subject of finishing outdoor railway buildings. While a proper modeller will want to achieve an effect that closely resembles the original building, what I want is something that will look nice and offer all-year-round weatherproofing. It is not clear to me that these goals can be integrated. If you look at an unpainted wooden building that has been exposed to the elements for fifty years or so, you'll see that the wood doesn't look like it does in Ikea furniture. In a clean environment, the wood will gradually darken, until it ends up a very dark brown, perhaps even black. If it is in direct sunlight this darkening will be evident even after a year. If the wood is in a dusty or dirty environment, particularly if it forms a floor, then the wood ends up a nasty grey-brown colour. Now, while a grey-brown shed may look entirely in keeping in an industrial rail yard, if I had a shed that looked like that in my garden, I'd paint it. The fact that a building is on a model railway does not, in my view, provide good grounds for placing a foot-square ugly grey lump in what is, after all, our back garden.Moreover, it seems to me that if you want to have a weathered wooden building on your railway, all you need to do is to slap a coat of wood preserver on the bare wood, and then leave it outside. It will weather soon enough, I imagine. The use of the wood preserver will prevent it weathering so quickly that it actually falls to pieces in the first winter outdoors. If this natural process isn't quick enough, you can get the weathered effect much more quickly by using coloured wood preservative. For example, Ronseal `Rustic Brown' wood preserver turns pine into almost exactly the same colour as it would have if it had been sitting around outside for a few years. What's more, it's dead cheap, easy to apply, and dries in minutes. Of course, if you want to simulate really old wood, which has been outside for a century or so, then you'll need to simulate the darkening effect of UV light. The standard technique for this is to mix india ink with isopropyl alchohol in a ratio about 1:50 (experiment to get the right strength). This very dilute solution makes a thin wash that is brushed onto the wood. The alchohol quickly evaporates and, unlike water-based washes, does not swell the wood. Alternatively, try black acrylic paint in white spirit for a really black, dull finish. Or, at least, that't the theory. The reality is that these are absolutely foul concoctions; the fumes sting your eyes, and Heaven help you if anyone strikes a match within 50 yards. What's more, unless you're going for the Scale Modeller of the Year competition, the effect of a water-based stain on wood of the appropriate size for G-scale is unlikely to be that dramatic. And even if you can get effective results, it's still important to bear in mind that india ink and acrylic are water-soluable, so you'll still have to waterproof the finished product because it offers no weather protection at all. I have been experimenting with applying blacking agents over a thick coat of wood preserver (judge the results below). The advantage, I hope, is that even if the blackening layer wears off, the wood will still be protected.
As an alternative to the india ink/isopropyl alchohol mixture, you could use one of the many wood dyes that are currently available from DIY shops. There's a huge difference between a wood dye, a wood stain, and a wood preserver, although manufacturers don't always use the terms consistently. By a wood dye, I mean a spirit-based, low-viscosity colourant. Wickes own-brand wood dyes are cheap and effective, and available in a range of colours. The `light oak' dye is a pleasant light grey-brown, while the `dark oak' is nearly black. Be careful: both these dyes lighten over time if not sealed. Because they have low viscosity, dyes don't need to be carefully brushed; the dye simply flows along the wood fibres. This has two benefits. First, it takes seconds to apply; second, it doesn't show brush strokes. Coloured wood preservers and lacquers are invariably viscous, requiring careful brushing to avoid irregularities.
The grimy jetty in the photo below was finished in two stages: first a couple of layers of coloured wood preservative, then a couple of washes of black acrylic paint dissolved in white spirit. The effect, I think, is reasonably realistic.
In the halt shown below, the platform is dyed with `light oak' and then sealed with a layer of matt varnish. The shelter has two coats of wood preserver, and has then been dyed the same way.
When using wood preservative, bear in mind that unlike lacquer-based finishes (see below), water-based preservatives do not form a film over the wood. This means that the preservative has to be worked into all the little nooks and crannies to avoid leaving somewhere for water to get in.
An alternative approach is to lacquer or varnish the wood, and go for something that looks nice rather than something that looks realistic. The advantage is that lacquers are virtually indestructible (although you'll need to get an outdoor variety for maximum strength). Lacquers are available clear or coloured; a coloured one gives some age to the wood and removes the Ikea look. This is the approach I have adopted on the depot at the top of this page. I am confident that this finish will withstand anything that a British winter can throw at it, and will look the same in ten years time as it does now. The problem is that it does look rather like a cukoo clock. Lacquer wood finishes, by their very nature, tend to smooth out variations in the surface texture. As a result, they result in a rather shiny, reflective finish. This is true even if you get a matt type. A partial solution is to wait for the lacquer to dry and them apply a thin coat of chemical paint preparation fluid (`Liquid Sand' or the like). This will roughen the surface at the microscopic level, and thus remove the shine. However, the result will never be as realistic as a wood dye.
Whichever finishing technique you use, consider whether you can usefully apply the finish before assembling the model. It can be extremely difficult to reach all the parts of the inside of a model when it is stuck together, but you can bet that the rain will find a way to reach them. My experience is that PVA glue continues to bond strongly to wood that has been treated with preservative or dye, but does not stick at all on wood that has been lacquered or varnished.
| Search |
| Shameless plug |
 By the author of this site. Buy on-line from Amazon USA | UK |
| Editorial |
So you want to be a university lecturer? Read this first! Speak like your boss: new developments in managerese |
| Computing features |
|
File handling in the Linux kernel: an in-depth look at how Linux handles files, filesystems, and file I/O
All sorts of Linux stuff Confused about CLASSPATH? answers are here First steps in EJB using jBoss (recently revised for jBoss 3.2) |