The Ghost In The Sleeper

Wooden railway sleepers don't last forever and often need replacing. The wood is of course still useable for other things. For example, it's not hard to find old sleepers used as fence posts. Such uses of old sleepers are easy to spot as, apart from their fairly standard size, you can usually see the impression left by the rail chair including holes where they were bolted down. Whilst you may see similar fences on model railways, how many modellers go to the trouble of modelling the ghost of the rail chair?

Whilst most people might not bother with such small details some people like to go the extra mile to make their models as lifelike as they can. One such person contacted me last year to ask if it would be possible to design a tool that would make it nice and easy to add these marks to 7mm to the foot scale sleepers.

The design work was relatively easy as I managed to track down a drawing for a rail chair (specifically a four bolt version used by the Lancashire & Yorkshire Railway) which I used to produce a 3D printed prototype on my resin printer. This was then tested using some of my son's Playdoh.

Of course the resin version is unlikely to hold up to repeated use being hammered into wooden sleepers, but having proved that the pattern it left seemed right, a little more work got me to a an easy to use tool that I could have 3D printed in brass.

Of course once the brass version arrived I couldn't resist giving it a go before posting it on to it's new home.

I think that looks pretty good, and as it takes just seconds to add the marks it's easy enough to quickly detail a large pile of sleepers ready to build a fence or maybe for an old siding where the rails have been removed. Given how many different ways old sleepers get used the possibilites are endless.

If you want to add rail chair details to your own 7mm scale model landscape then I've made the tool available to buy on Shapeways (see here).

While four bolt chairs were fairly common, I do have drawings for rail chairs from other companies which (given time) I'm intending to use to produce a set of tools, but in the meantime if you want a tool to produce the ghost of a specific comapnies rail chair let me know and I'll do my best to help.

Hinge Versus Flex

When I built my first working point at the beginning of December I mentioned that I was intending to build another one to see if relying on the rail flexing was a better approach than hinging the rails. I actually built the second point quite soon after but for some reason didn't post about it. Anyway here we have both points next to each other with the new version no the left.

So how does allowing the rail to flex compare to using hinges? Firstly aligning the rails accurately and getting a nice smooth curve through the point was much easier the second time around. Partly this was down to having a better idea what I was doing when shaping the rails but also as I could use the rail gauges to position and hold the rails in place while they were soldered on which was easier than holding them in place and drilling holes for the hinges. On the down side though I'm honestly not convinced by the motion of the point. Firstly as the rails are flexing I've ended up with a point that when left alone sits with the tie bar half way between the two routes. Secondly keeping the point one way or the other requires quite a considerable amount of force.

So in conclusion I'm going to stick with fitting hinges to the points and hope that with more practice the flow of the rail will improve slightly over the first attempt so I'll have both nice looking points and points that require little effort to operate.

Rail Cutting Jigs

So having given you a few days to think about the mystery 3D printed objects I'll put you all out of your misery. They are jigs to help cut rail at the correct angle for making Hudson Type 1 and Type 2 points in O14. While I have the KBscale jigs for actually soldering the rails together they rely on you cutting or filing the rail ends to either 15 or 9 degrees and I found doing this accurately was a bit of a nightmare and so a jig or two sounded like a good option.

The printed jigs are actually the third iteration of the basic idea which was to trap the rail in such a way that a razor saw could cut it at the given angle. The first attempt, which I've unfortunately thrown away so can't show you, used four pieces of 4mm square section styrene stuck to flat sheet, two on each side of the rail. The pieces on each side were offset so that you ran the saw across the rail resting it against the ends of the blocks. It worked well enough to produce the first crossing I made but it wasn't easy to hold the rail still while pushing the saw against the blocks.

The second jig used 12BA nuts and bolts to both clamp the rail and to help position the saw. The screw holes being carefully positioned so that they both allowed the nuts to grip the web of the rail, but also to act as the guide for the saw. While this was much better at keeping the rail in place it was still difficult to keep the saw in place against the upright screws, there was a danger of sawing into the nuts, and the cheesehead screws meant that the jig wasn't very stable on the worktop. It did, however, allow me to build two Type 2 point without requiring any extra filing of the rail ends.

The printed versions continue the idea but without relying on the screws to guide the saw. The two printed parts are used to clamp the rail in combination with grooves in both halves. The slightly bendy nature of the strong and flexible plastic is actually really useful here as it allows you to really tighten the jig so that the rail doesn't move at all. The saw then fits within in the guide formed between the top part and the block on the bottom half making it much easier to cut the rail.

I haven't had a chance to build a point using the new jigs but I have cut a couple of rails to check that they work and they appear to work very nicely which should make it easy to churn out more crossings in the future.

Turn to the Right

Having now recieved my order for more rail I've continued on with my O14 point building experiment. The first attempt was for a very tight Type 1 point which in retrospect was going to be a struggle even for short wheelbase locos so for my second attempt I've switched to the slightly less severe Type 2 point. Of course with more rail to hand I've gone a lot further than just building the central frog section, in fact I've built an entire point.

I've deviated from the suggested instructions in quite a few places though as I've built the point onto copperclad sleepers rather than the suggested plastic sleepers. This meant that I had to gap the rails to ensure that the polarity of the frog could switch when the points changed. I also simplified the tiebar slightly by not trying to fit nice looking plates and rivets but using a bit of wire instead. Of course photos on their own don't show you if any of this was successful or not so here is a video of me testing it.

So yes it does work, but I'm not entirely happy with it. Mostly I'm not happy with the hinges. Drilling the holes to make the two parts of the rail line up properly was a real pain. So having recently seen someone else build one without hinges where they relied on the rail flexing instead I'm going to have another go and see if that is a better approach. Still not bad for a first attempt.

The Makings of a Point

I've really enjoyed working in O14 building the Clayton loco, but if I want somewhere to run the loco that isn't a simple straight bit of track I'm going to need to build some points. When I was at ExpoNG I picked up both the Type 1 and Type 2 crossing jigs from KBscale in preparation for this but it took until this weekend to find time to have a go. Even with a jig to help hold all the parts it's still quite tricky (or maybe I just haven't had enough practice). The first problem is finding a reliable way of cutting/filing the ends of the rails at 15 degrees. In the end I built a small jig from styrene which seems to do the job but it feels like there should be an easier way. Once all the rails have been cut and fitted to the jig the problems don't end. Because the jog is metal it's impossible to get enough heat into the join with my soldering iron as it just soaks into the jig instead. The trick is to use the kitchen blow torch (usually used for creme brulee) to pump in lots of heat.

Once it's cooled and cleaned up it looks as if it will do the job nicely. Unfortunately I then cut the rails back but measured badly so they are too short for the type 1 turnout I was trying to build. Not to worry though as the practice was well worthwhile. Unfortunately I seem to be almost out of the PECO IL-115 rail so will have to order some more before I can have another go. Also looking at the instructions I'm not sure I like the suggestions for hinging the switch rails (a wire down through the sleeper). I'm wondering if I can just do without a hinge and rely on the slight flex of the rail, but that experiment will have to wait until I get some more rail.

Now With Sleepers

I've spent the last couple of days slowly adding the sleepers to the rails so I can see how the section over the bridge will look. The sleepers in the undergrowth can wait until later.

It's not a perfect replica, when compared to Bobby Mhor's photo, but I'm pretty happy with how that looks.

Rocks and Rails

Just a quick update on the diorama, as I've now scribbed the stonework on the right-hand bridge support and cut the rails to length.

The rocks were scribbed into the plaster using a dental pick, which worked nicely. The pattern isn't fully accurate to the actual bridge but I think it should work reasonably well once painted up and part hidden under plants.

The ends of the rails are likely to be hidden where the track is over grown so I've soldered them to copperclad PCB strip for strength rather than relying on the hand spiked sleepers to hold everything in place. This should also make fitting the sleepers easier as the gauge is already set.

While neither of these tasks are huge in themselves, I do feel that the scene is starting to come together nicely now. I see I still have a bit of work to do where the girders join the left bank (that gap needs filling) before I start painting the rocks, but I can now also crack on with building the track and detailing the girders and those jobs should move the diorama forward quite quickly.

Dual Gauge Track

While most of my modelling recently has been focused on my entry for the Dave Brewer challenge in October I've not forgotten about the Clayton battery locomotive.

Mostly I've been writing a long list of all the small issues I've found with the current design now I've got the first prototype print in my hands. It's turning into quite a long list, but fortunately most of them can be solved on the current print with a little bit of thought, which allows me to test the changes before ordering another print. Before I can assemble and test the locomotive though I need some track to run it on.

If you remember I'm building this model in O14, 7mm to the foot scale running on 14mm track to represent 2 foot gauge track, as the loco is too small to model in my usual scale of OO9. The problem is I've never modelled in O14 so I didn't have any 14mm track. Fortunately I'd planned in advance for this and ordered some roller gauges from KBscale. Given that I'm short on desk space I decided to produce a test track that would let me test both O14 and OO9 (O9 and Gnine as well if I wanted to branch out) models, so I now have a 60cm dual gauge test track.

OO8: 4mm Scale, 2ft Gauge Track

One of the problems of modelling in OO9 (9mm gauge track in 4mm scale) is that often the track isn't really to scale. 9mm track at 4mm to the foot is 2' 3" but narrow gauge railways varied quite widely in their choice of track gauge. In Britain possibly the most common gauge was 2 ft meaning that 9mm track is 1mm too wide.

The bridge I'm building for my entry to this years Dave Brewer challenge carried a 2ft gauge railway and given I'm aiming for the smallest entry category (to make it easier to transport on the train) I thought it made sense to produce track to the correct gauge so I give you my first attempt at OO8 track (8mm gauge track modelled at 4mm to the foot).

As you can see this combines the numerous experiments I've been blogging about recently (here, here, and here) to produce a short test piece. Producing actual track using rail spikes in this scale is exceedingly time consuming. I've probably spent about 10 hours producing this small piece, although by the end of that I'd got a system going that means producing more will be quicker but not by much. On the plus side the track is to scale and an accurate representation of the real track I'm modelling right down to the number and placement of rail spikes. The spikes look a bit too big in the photos and they probably are, but in real life they blend in quite well.

I'm really happy with how the track has turned out, although it has meant that the number of hours I thought I'd need to build the diorama for the challenge has increased quite substantially. I should still have enough time though so I'm not worried.... yet.

Rail Spikes

So, having arrived at a much better approach to painting rusty surfaces I've now turned my attention to the track, specifically how I'm going to affix the rail to the sleepers. The bridge I'm using as inspiration for my entry to the Dave Brewer challenge, which is still a secret for now, uses a fairly light rail held to the sleepers using metal spikes.

All the track I've hand built before has used copper clad sleepers and I've simply soldered the rail in place. Clearly this won't work for rail held down by spikes. So firstly I'm going to use sleepers made from plywood (hence my experiments at painting wooden sleepers) but more importantly I need a source of spikes that will hold the rail in place. My intention is to use code 40 rail, which is pretty small, and I haven't been able to find anyone who sells spikes for such small rail. What I did find though was a series of articles in NG&IRM Review by John Clutterbuck (which were republished in 2008 as a free download) about making track. While John models in 7mm scale using larger rail, he happens to make his own spikes from..... staples.

The approach I've taken mostly follows John's description but with a few changes to take into account the smaller rail size.

I bought the smallest staples I could find but they are still made from quite a thick metal (about 0.45mm) so having used masking tape to keep them together I filed the edges down a bit to make the corner a lot thinner (I had to use diamond files to really make much impression on the steel). Once I'd filed down both sides I used a set of track cutters to cut off the legs of the staples keeping just a tiny amount of the curve. All the spikes were then separated from each other before soaking for about 5 minutes in nail varnish remover (the stuff I "borrowed" from my wife is acetone free but I'm guessing acetone would also work) to ensure the metal is clean and that the plastic coating applied to staples has come off; you end up being able to see the bits of plastic floating in the solution. The spikes were removed from the nail varnish remover (a magnet makes this nice and easy) and dried before being soaked in Carr's metal black. They turn black almost instantly if they are nice and clean but again I left them for about 5 minutes, before removing and drying them.

I still need to figure out the best way of actually using the spikes to fix rail to sleepers but for a quick test I drilled a small (0.45mm) hole and pushed a spike through the sleeper, added the rail so I could accurately drill a second hole for the second spike. Once both spikes were firmly pushed through I held them in place with a drop of superglue added from below. Once this had dried I cut off the excess spike and filed them flush with the bottom of the sleeper.

For a first attempt at making and using rail spikes I think this has worked well.

Painted Track

So having settled on an approach to painting sleepers I spent today (on and off while doing other things) finishing off the hand built track for Hudson's Pizza. First job was to fill the gaps in the sleepers with milliput. Pressing a small amount into the gap on each sleeper smoothing it down and removing the excess takes a lot longer than you might imagine and then of course you have to wait for it to harden.

Painting was slightly more fiddly than in the experiments I did because of course the rails are soldered in place and blackened so I needed to keep the paint off them. First up was the white gesso, then a thin wash of RailMatch sleeper grime (50/50 paint and water). I then darkened the sleepers using a black wash which while bring out the details did make things a little too black. This was fixed by dry brushing a little sleeper grime and then a little Model Color brown sand to give a result that I'm really quite happy with.

Chemically Blackened

A very short update on the track building today. I spent about ten minutes cleaning the track with a fibreglass pen before chemically blackening the rails and then burnishing the rail top with a wooden coffee stirrer to remove some of the blackening and ensure a good surface for locomotives to collect power. Everything still works and it now looks a bit more prototypical. The sleepers will be next.

Sleeper Painting: Experiment Number 3

Having thought about the sleepers a little more I realised that the main problem with the original approach wasn't so much the colour as it was the texture; they were horribly smooth and unnatural looking. After a few suggestions from different people I've come up with what I think is a simple yet effective approach. The sleeper shown here (the furthest right) is painted using the same basic colour as the other sleepers (RailMatch sleeper grime) but with some additional texture.

The trick was to prime the sleepers using some white gesso which provides a textured surface for the paint to stick to. I then thinned the sleeper grime down a little so that it wouldn't obliterate the texture and finished with a quick black wash. This seems to work well giving both texture and some colour variation and is remarkably easy to apply as you don't need to be particularly neat about the painting. With a little extra weathering as the track is bedded into the layout I think it will work well.

Equal Opportunity Modelling

As an equal opportunity modeller I thought I'd allow my other locomotive to experience the new track which gives those who prefer steam locomotives something to watch.

Round and Round She Goes, and Where She Stops Nobody Knows

The last time you saw any work on my new layout there was just a new paper track template. Things have moved on quite a bit since then.

Firstly I spent quite a lot of time at the mind numbing task of cutting and then filing a groove into the 90 sleepers required for the layout. I used the same jig as before to help measure and cut the sleepers which definitely makes the job easier although it's still rather boring. The sleepers were then all attached to the template using double sided sticky tape. That is the easy part, next up is the rails.

Obviously the rails needed to be curved into shape before I could try soldering them to the sleepers. Curving them as I went would just have ended in disaster. Now I don't know of anyone who sells a rail bender for OO9 gauge but I bought one from KBscale designed for use with a slightly large rail (their market is 7mm scale modelling). This consists of three sets of washers bolted to a plate. Sliding the rail between the washers introduces a curve into the rail while keeping it flat and not allowing it to twist. I used the tightest setting and it worked perfectly. In fact it produced rails with a curve slightly tighter than I needed.

With the rail nicely curved I soldered on the outside rail first using the template to guide the positioning. I needed two pieces of rail to complete the circle so there are two gaps, one is a nice flush join, the other is a little wider than I would like, but I've now filled it with some solder and filed it back and it seems okay. With the outer rail in place I then soldered the inner rail in place using the excellent rail gauges Paul turned for me. Again two pieces of rail were needed to make the circle but I made a better job of the joints.

Of course having two rails at roughly the right distance apart doesn't mean they will work as track. The only way to know for certain is to add a locomotive and some electricity and see what happens...

I thought I might have to adjust a few tight spots but I couldn't spot any obvious problems so it looks like the track is done (structurally at least, it still needs painting etc.). The video was taken before I filled the over large gap in the outside rail which is what you can hear as the loco passes through the left side of the circle. With the gap filled in the noise is a lot less obvious.

While I was really happy with the short test track, given it was my first attempt at building track, I'm really impressed with how this has come out. I always assumed I'd just be buying track when I built a new layout so I owe a huge thanks to Paul for providing the gauges and everyone who has provided suggestions, advice, and encouragement!

Sleeper Painting: Experiment Number 2

Okay that didn't take long. Here we have a rough experiment of painting a metal sleeper. This again follows the same approach I use on the metal parts of the wooden bodied wagons and I've shown some rust where the rails meet the sleepers. I've also included a photo of both approaches for a comparison.

Any thoughts?

Sleeper Painting: Experiment Number 1

I'm still trying to clear enough space on my desk to get cracking on Hudson's Pizza but I have now cut and gapped all 90 sleepers I will need for the circle of track.

While I was really happy with how my first attempt at hand built track turned out I've been wondering if I can make it look better and slightly more natural. Specifically I was thinking it might be possible to improve the sleepers. On the test piece I simply painted them with RailMatch sleeper grime. There were two problems with this approach. Firstly there was no texture to the sleeper tops and secondly the colour was very uniform.

What you can see on the left is my first attempt at trying something different. The sleeper isn't soldered to the test track just placed underneath for the photo. Basically I ran a file along the sleeper to add some wood grain and then painted it in roughly the same way I do my small wooden bodied wagons albeit with an extra round of dry brushing using Model Color London Grey to give a slightly more silvery appearance. If I'd painted this in situ then there would have been a bit more variation as it would have been difficult to dry brush right up against the rails. Now I'm not sure I'm entirely happy with the final colour but I'm also finding it difficult to work out if I prefer this to the block colour of the previous attempt. Any thoughts or suggestions?

And before anyone says anything, yes I know Hudson's would likely have used their own metal sleeper track system for the test track, but I'm only basing my layout on it and I think wooden sleepers will be easier to model.... possibly. Maybe I should have a go at doing a metal sleeper?

Pizza Track Template: Take 2

Another day, another track template. After posting about the track template yesterday I took another look at it and realised I'd messed up. The template consists of four pieces of paper stuck together and I initially thought that the white space at the edges of the sheet was intentional so cut them off before joining the sections to give a single template. Unfortunately that wasn't the case and cutting of the paper meant the circle wasn't completely circular. You can see this in the join at the front in the photo from yesterday.

I also decided that I'd probably got the sleepers a little too close together on the original template, so rather than just reprinting I've generated a new template with only 90 sleepers (down on the 120 sleepers on yesterday's attempt). I think this looks better and given how tedious and time consuming cutting the sleepers is I think it's a good decision.

Pizza Track Template

As my modelling area starts to reappear from under the detritus of Christmas I've made a small start on my new Hudons's Pizza layout; I've designed and printed the track template.

The next job will be to cut the 120 sleepers required for the full circle before I can think of doing any actual track laying. This time I'll also remember to gap the sleepers before soldering on the rails, especially as it will be much harder to do afterwards on a circle of track than it was on the straight test piece I built before.

Go Your Own (Permanent) Way

As Paul was kind enough to turn me a set of brass track gauges I no longer have any excuse for not trying to hand build some OO9 gauge track. Paul has already used his gauges to build a working point, but I decided to start with something a little simpler; a 30cm straight piece of track.

Of course, deciding to have a go at making my own track was the easy part, working out which parts I was going to use was a bit harder. The easy part was the sleepers. I'm using 3mm wide strips of copper clad PCB cut to a length of approximately 19.25mm (based on a couple of measurements from some PECO stuff I had kicking around). The rail was a lot harder though. The OO9 gauge track I've bought previously uses code 80 rail. The rails are measured in 1/1000's of an inch, so code 80 is 0.080" high. For a light railway this looks a little on the large side and many people use a lighter rail. For example, Paul used code 55 steel rail to build his point, which is 0.055" high. The problem I had was that I wanted to avoid using steel rail as previous experience shows that it does corrode over time, much more so than modern nickel silver rail. In the end I've compromised and used code 60 nickel silver rail from PECO. This rail (PECO IL-1) is a little odd in that it is designed to be used for either Z gauge (which is really tiny) or as conductor rail in 4mm, but it seems perfect for OO9. C&L Finescale stock both the sleepers and the rail and so an order was placed.

Once the parts turned up the first job was to cut the sleepers. The copper clad PCB turns up in strips the right width (3mm) but 17" in length. Fortunately a razor saw cuts through the PCB easily, the problem is how to cut consistent lengths. I could have just measured the strip and marked it up, but I decided I'd make up a little jig to aid the process. This jig sits nicely in my small mitre box and when pushed up against the blade of the razor saw ensures a consistent sleeper length. Once I'd cut a couple of strips into separate sleepers, I then turned to building the track.

I started by drawing up a template on the computer to aid in positioning the sleepers and rails. This was easy for a straight section, but if want to build more complex track in the future will need a little more thought. The sleepers were then stuck to the template using double sided tape, and then the first rail was soldered in place, using coffee stirrers to hold it down while I applied the soldering iron. Once the first rail was soldered to every sleeper I could move on to the tricky bit of adding the second rail.

It turned out that with the track gauges Paul had made me, and some more coffee stirrers, fitting the second rail was nice and straightforward and before long I had something that actually looked like a piece of track. Even better a quick test with a wagon and it looked like it might actually work too. What I'd forgotten to do though was to file a gap into each sleeper to ensure that the rails weren't electrically joined. Fortunately on the straight track it was fairly easy to get a file in between the rails to remove some of the copper; in the future I'll definitely be doing this before I solder the rails down. Finally I removed the track from the template, soldered a wire to each rail and then had a go with letting an actual locomotive run along the track.

I opted to try it out with the small diesel I built from the Narrow Planet kit, as the motor would be a lot easier to replace if everything went wrong than it would on the Quary Hunslet, but fortunately, as you can see, it all worked perfectly. In it's current form the track isn't going to win any beauty contests but it should clean up nicely and now I know that I can at least build plain track which is useable so I'll call that success.