On Saturday I visited the Great Electric Train Show (GETS) 2024 hosted by Key Publishing and Hornby Magazine at the Marshall Arena in Milton Keynes.
I’ve not been to this show before and I really enjoyed it!
The highlight for me was being remembered of my time working on the redevelopment of the Catesby Tunnel complex as I came across the OO gauge model of Catesby Tunnel and nearby Charwelton station by the CBM Diesel Group. It’s so crazy to see this as it spans the whole development area and is laid out so accurately that I recognised it from the model (not being aware it was actually at GETS).
I remember showing the Stepnell team pictures of this model when it was in BRM a few years ago. Coincidentally, it is on the cover of the issue with my first model article in it!
I also remember commissioning some bat surveys on the road bridge just before the tunnel too.
I had a look at a few new items due to come out soon. The Cavalex Models Class 60 is looking great, as is the Heljan LNER Garratt (that will look excellent on my garden railway). Hornby’s Locomotion No.1 is also fascinating to watch in motion.
Here is a lengthy video of mine from Day 1 of the show:
There were some really great models on display and I got to catch up with some friends and have a quick wander around Milton Keynes too.
Having plucked up the courage to lay the long straight of track including wiring it and pinning it down, my next task was to work out how to curve the track, for the many long radius curves I’ve designed into the garden railway.
The first step would have been double-checking all the main radii of the corners designed into the super-structure, but unfortunately for me, I seem to have misplaced a load of photos and screenshots from my phone dating back to when I switched phones, so I don’t have all the many screenshots I took of my kerfing calculations. However, I did take rough notes and I was fairly consistent throughout so I should be able to manage.
Returning to those kerfing calculations, it is important to remember that the radii I calculated apply to the inner circle of the kerfed wood, and although I mostly calculated both an inner and an outer radius for each section of the platform, technically the difference between them is the width of the railway top (274mm) minus one width of side frame (22mm). This is because the outer curve calculation is still based on its inner face, which is the inside of the side frame (thus removing 22mm). So the true distance between the curves is 252mm. I seem to have fuddled this a little on some parts of the railway, but no one will notice.
The inner OO track is positioned approx. 4cm from the inner edge of the railway top, so it is safe to assume I need to add approx 1.5 inches to the inner radius to get the correct track curve. The first curve I am working on is the tightest and most challenging, which is the one leading into the shed. I’ve decided to do this one first as it is the one that would be best to complete now so I can add a roof to my tunnel and seal the shed. This meant that the OO track needed to be approx 33.5-inch radius, which led to the next challenge: There are no Tracksetta templates available for some of these larger curves.
I decided to create a radius template on some of the offcuts of my ply railway tops, which I did by pinning a track pin at the base of the length and then marking out the main ranges of curves used on Grass Heart Vale and then I used a metre ruler like a massive compass by resting the hole in the ruler over the pin and drawing the radius along my measured out marker. The largest curve on the railway is approx 66 inches, so I used a second length of ply to extend this and drilled a hole in 1.8m length of aluminium bracket to do the same and made a super compass! This gave me two boards with all the key radii measured out acting like a radii ruler thus also allowing me to do measurements between the marked out lines.
I then set the track up on the 33.5-inch template and gently curved it to match the stenciled outline and used these Proses rail curving tools to hold the track to the correct curve. They come in packs of 2 but I have 4 of them to help me do my longer curves. I then set up the track in the tunnel and started pinning it down, but leaving some flex at either end of the track sections to connect it to the rest of the line.
To get through this section of line I wanted to make sure most of the curve is complete in one rail section to limit any risk of rail joiners causing a ‘dog leg’ in the track. Near the entry to the shed and at the end of the rail section, it seemed that I would need to tighten the corner down to about 30 inches to make sure it could cross into the shed at the desired point and thus meant I needed to connect a new track section and track joiners on a curve.
I limited the risk of having a ‘dog-leg’ rail join by staggering the rail joins (rather than having them both parallel) and moved the one rail join back by about 6 sleepers. This is slightly fiddly as you have to move the rails in and out on two sections to match and account for where the rail joiners will slide over sleepers. It also means I can’t solder up the track while it is loose, but in fairness, I couldn’t do that on curves anyway. This allows expansion to take place but limits the amount of tension on two break points and instead spreads that tension out across the rails better.
Once all that was done I then connected the straight sections I laid last time to the tunnel curve. I’ve read a lot about ‘easement curves’, but I don’t find many youtube tutorials explaining them to be that well communicated. They talk about the ‘point of tangency’ as if you already know exactly what that is and then discuss an offset on the straight, but, again, they don’t really explain what any of that means for laying the track. Here are a few anyway (maybe you can make more sense of them than me)
I found a more helpful graphic on easement curves and experimented with it at the tunnel face, although I think I messed it up a bit as I calculated the offset on the wrong side of the curve!!! It didn’t really amount to much of an error as the curve transition including the straight is about a 20inch length and found an organic easement of its own, but required releasing a few track screws I’d placed on the curve to improve its shape.
Doing the parallel line of OO gauge track was much easier as I used the other Proses tool for laying the parallel track and that was simply a matter of laying the track matched to the inner radius using the tool (with a bit of assistance from the curving tools).
I tested the curves with the longest wheelbase loco I think I have, which is the Hornby P2. Although it is indicated that it can go around 2nd radius/17-inch radius curves, I’ve seen it struggle on my 21-inch curves on the carpet railway and occasionally derail itself, so I thought it would be the best loco to test with its x4 flanged axles. I also tested the curves with my recently purchased Hornby 9f, which is a longer wheelbase 2-10-0, but the middle driver is flangeless, so perhaps better at curves than the P2. Anyway… regardless, both test locos ran around the curves with no problems at all.
I then took my learnings and started work on the longer radius curve leading towards the front of the house and used all the same techniques. Apart from some grade variation between the platform sections all the track runs very smoothly with no issues. In fact, I think I did a much better job at calculating the easement curve this time, but again, I don’t think it needs that much overthinking at the lengths of track I’m working with as it all ‘eases’ into itself. I do occasionally find the Proses curve tools don’t create an entirely smooth curve, particularly the bits of track between each individual tool, but I don’t think anyone will really notice and I am being quite pedantic, really. The Tracksetta templates provide a better curve (if you can use them), basically.
I also plan to experiment with super elevation on the rest of the curves, but I can return to that once the track is down as I can unscrew the sections to account for that, but importantly… so far so good.
I started work on my Grass Heart Vale Garden Railway in May 2023, and as of the start of August 2024, there is finally track being laid down onto the railway tops.
I must admit I was quite nervous about moving onto this step, having spent so long learning how to make curves in the superstructure, designing bridges, and completely upgrading the interior of the shed with lights and power, I felt I hadn’t made much space to learn how best to lay track.
The last time I did this was quite a long time ago now for Brewery Pit and I didn’t really give it loads of thought at the time, other than making sure the layout could operate under DCC power.
My original plan was to wire a bus cable under the entire length of the run and feed the wires up to track sections. I would never rely on rail joiners alone to transfer power across sections of track, but the added complications of metal expansion and contraction of the rails in the varying UK temperature meant I had to ensure this was accounted for. Laying out a bus wire felt like it could be quite labor-intensive and I was nervous about cutting holes in the wood I’d just spent a long time sealing from the rain. I saw that some OO garden railway makers had instead used the track as the main bus wire, by including loops of wire soldered between each track section (around the rail joiners). This provides electrical continuity, and leaving some excess cable at each track section also allows the rails to expand and contract. I thought about this method for quite some time and the pros, at least in the short term outweigh the cons. In my mind the wiring method is exactly the same as what I would be doing with a bus wire, but on the surface instead of beneath. The only negative I can imagine from this is whether you would achieve a more consistent current via a larger bus beneath than what you might get via passing current through smaller gauge cable connecting rail sections. If this turns out to be the case I can always install a bus cable later and it will mean that I don’t have to bring connecting cables up through the boards as frequently.
I purchased cable that is fairly new to me for the job of soldering these section loops. it is PTFE Silver Plated Copper Equipment wire which is apparently used in electrical equipment as able to tolerate high temperatures and is resistant to various residues. I bought a few different gauges to test and settled on 24 AWG for the OO track and 22 AWG for the 16mm track. Unfortunately, I could only get the 22 AWG cable in white, which is a bit obvious on the track surface but we will see whether I can sort that out at a later date (edit – I have just ordered some alternative black cable to replace it with).
I then started trying to work out the final spacing of each track line and I purchased some tools to help with this made by Proses which help align parallel lines and also help hold consistent curves. I also have various Tracksetta gauges, but most of my curves are wider than the templates allow.
I also purchased a lithium battery-powered Dremel multi-tool, as I thought that would be a great help when it comes to drilling holes and cutting track.
I decided that the sensible thing to do was to focus first on laying out the track on my original test section which is a long straight. I thought that would help me get into the groove before I start trying to work on curves. I started by working out how close I was willing to get to the entry of a curve (18 inches) and then laying out the straight sections. I installed the rail joiners coated in vaseline (to avoid them getting bunged up with solder) and then soldered up the silver-plated cable. I used my heavy-duty 80W soldering iron for this with a 6mm tip. I first found it a bit unruly for the task, but the size of the tip actually prevented me from making a mess of the track. I also quickly learned that it is best to cut the last three sleepers away (connected together) to allow space for the rail joiners and the soldering exercise (at least on straights).
I did all this for the OO and 16mm track in situ and then gave some thought to how I’d fix it all down.
I recently came across the existence of track screws (rather than track pins), which I thoughts would be very helpful and allow some options for maintenance of the track later so I decided to use these to fix the rack down. I also screwed them down into the sleepers with parallel crews on the outside of the rails. This should limit the rotation and slipping of sleepers when the rails expand and should hopefully mean the expansion just moves along the set path of the track. To account for this expansion I’m leaving a small 1-2mm gap at each rail joint. Apparently, if you lay the track on a hot day you can butt them up to each other and they will retract later, but I didn’t find the surface temperature at the railheads to be that warm (also: see various comments about summers in the UK), so I have left a small gap regardless.
The Dremel purchase was worthwhile in the end as there is a certain surgical precision required to make pilot holes in the ends of sleepers, but it all worked quite easily. My only concern was that I wasn’t sure what metal the PIKO track screws were made from and how quick (or if) they would rust, so I’ve sealed them with waterproof PVA for now. including the surrounds of the where they enter the board.
Once all that was done it was finally time to do some tests with some locos and all is working well so far, but we will have to wait and see how the weather treats my efforts.
It’s come to my attention in the past few days that I wrote very little about building a commission of Thunderbird 3 for one of my friends.
He had built one himself from a rather nice resin kit but wasn’t happy with the finish so he asked if I would make him a version close to the props used in the show. I enthusiastically said yes, as I was keen to branch Grasslands Models out into commissioned builds and had recently taken on a chassis build for someone else.
I did a fair amount of background research into Thunderbird 3 props and the different sizes of props. The resin model of Thunderbird 3 seemed to be based on some larger-scale props and the panel lining seemed to trace its origins to some of the other props. I was aware that when the BBC revived the series in the 90s, there was a series of new promo shots made with purpose-built models for use on comics and posters and stuff. Quite a few of these props were made by Martin Bower and it was these props I used as references to match the pain scheme.
Like many model-making escapades, the final livery/colour schemes are a hot debate and TB3 is always one that is debated a lot on the correct hue of red. It is near orange in the final produced show, but in the two films and the promo shots mentioned above it is often a bit more red. You will also find very few pre-made toys or models that reflect TB3 in anything other than a red (rather than an orange), so I settled on matching the more red colour scheme. I did some test flashes of various Halfords spray can reds and asked for input from the customer on his preference and we settled on Rover Vermillion.
Now, TB3 was less a build and more a paint job, really, and a carefully staged one at that! I first primed the main sections grey and then primed the rims of the rockets white and masked them off before an overall second prime in red and then the overall main livery application of the Halfords red. I then did the retro rocket ring in Ford Nimbus Grey.
The next section beneath the retro ring has a fairly complex colour scheme which I spent a while trying to work out how to do. it’s a navy blue section with red stripes which a series of grey fins sit on. Research seemed to show that the blue was originally achieved with blue insulation tape, which I think I did consider or experiment with at the time, but I wasn’t happy with the result so I used a blue spray can instead. I then used fine masking film carefully cut to specific widths and masked them over the blue to create the red stripes.
After painting up the fins I then carefully went about fitting them to the fuselage.
Most of the parts were fixed with gorilla glue, as gorilla glue is resin-based it sets on the resin parts really solidly, however, gorilla glue expands so you have to be very cautious with how much you use, so there was lots of post-gluing finishing I did. I also pinned the majority of the parts.
After adding some other details I then gave the model a high gloss coat for the transfers to settle on. and then very, very carefully went about fitting them as accurately as possible to the model.
I then protected them with another glass coat. I finished off with some weathering and decided to go for three different grades of weathering on each of its sides, so a light weathered side, medium weather side and a heavy weathered side.
With the model all finished I took some final photos for my records and sent it off to my friend who was very happy with it and I occasionally get updates on where it sits in his house.
July has certainly been a strange and quite intense month. Work has been very busy with lots of my colleagues away so I’ve been overseeing things and I’ve been taking a surprisingly long time to recover from a bout of COVID picked up last month. Lots of members of my family have also injured themselves or are recovering from operations. That’s a fair amount of stress in the old body system! Through all of this focussing on upgrading the shed has aided keeping my faculties in the right place, but it has led to a strange period of, now 18 days, of me working for a few hours in the shed to sort things out each day. The weather has also been challenging on that front as the whole garden project was really planned for sunny days and there have been precious few since I started this project back in May 2023.
Since my last update on the shed where I fitted the ventilation and created some support beams for the rock wool insulation, I’ve completed all the rock wool insulation installation, fixed the ring main after I unhelpfully shot a staple through the cable and broke the earth core in two, which meant rewiring about a third of the shed (the trickiest bit too).
I also installed a moisture membrane, which was actually much easier than I expected it to be, the thickness of the PVC sheet really helps as it folds and settles quite well as you slowly fix it around the walls. I used a staple gun to fix in place (thus the ring main issue).
I then measured up the inside of the shed so I could purchase the final wooden surface. I decided to use OSB board as it is fairly cheap and rigid. I opted for installing 18mm thick OSB on the back and side walls with a 9mm thick bit of OSB on the front of the shed beneath and above the windows. Measuring up the correct main cuts for these boards and then fitting the 2.4m lengths in the car was a particularly stressful day and involved a very careful drive home, constantly worrying that my harnessed-up payload of wooden panels would surprise cars behind as it excitedly leaps from my boot all over someone’s bonnet, but rejoice, I got home safely (albeit slowly).
I’d given a fair amount of consideration to the staging of the wall installation. I decided I needed to allow a lip at both the bottom of the walls (to slip a PVC workshop floor underneath) and also at the top (to allow me to rest the ceiling panel on and aid me fixing them up on my own). The main challenge with these boards was making holes for the vents and power sockets, which in the end was probably easier and tidier than expected. The main mistake I had made was installing plastic tubes to the sides of the power sockets which would really get in the way of the walls. I removed these in the end to allow the walls to fall flusher and closer to the sockets. I didn’t mess with the ones around the consumer unit as that seemed like a recipe for trouble and instead designed a raised box to surround the fuse box with. The other thing I learned from this process was that although my woodcuts are straight, the shed certainly isn’t! Therefore quite a bit of time was spent calculating slight changes in angles to fix the walls properly. It all took quite a long time, but it all went up well. The most challenging was the middle width of 18mm OSB on the back wall as it was a heavy bit of wood and nearly 2.4m long. I attempted to test fit it and once I’d lifted it and got it into a very precarious position I decided there was no way in hell I was taking it back off so I made the adjustments to it in situ.
Having learned a bit from the handling of the larger bits of wood on the back wall, I decided to cut the ceiling panels into thinner strips that covered the length of the shed. Fitting these was fairly easy, even while on my own as I used a set of extendable ladders to wedge the far end on while I screwed the panels into place. The lip I’d left also helped with the front and back panels too.
Although I first tried painting the OSB board with white wood paint, I found it too thin and was taking a long time to get decent coverage. I then did some research and found that masonry paint was the preferred paint for OSB board and I found it much thicker and able to get coverage after a few coats. I did the first two coats with a roller and then a final coat with a masonry paintbrush.
I also started doing some work on the exterior of the shed as it has suffered quite a lot in the very wet winter of 23/24. So I applied some wood rot treatment, some wood hardener, and some layers of waterproof PVA and wood filler to the areas around the windows. I still have some work to do here, but I prioritised getting this done in the few moments of sunshine.
I then returned to the railway itself and the support frame for this section of the track bed that I installed last year. I first painted the viewable area in the same scheme as the rest of the railway and then purchased my final sections of 18mm marine ply for the running surface and cut it to shape to fit this remaining section of the railway.
At this point, it seemed wise to open up the portals for the railway to enter/exit the shed and I cut them to their final width, which I felt more comfortable about now that I have purchased the parts for my shutters for the portals.
The rain makes this part of the job challenging as I need to install, treat, waterproof, and lay the roof felt surface onto the railway tops in one day as any rain can mess up the process while I’m in the middle of this job. Yesterday on a sunny afternoon after work, I managed to do the final coat of paint and PVA and cut out the roofing felt shape of the entire final section of the railway top.
After widening the railway portals to their final 7.5-inch height, I went about fixing the railway tops and then carefully moved the shaped roofing felt into place and it all seemed to go together surprisingly well!!! I fixed one end of the roofing felt down with clout nails, but will tidy up the rest later. I’ll also install a border at the rear to tidy up the transition between the walls and the roofing felt.
So there you have it, nearly 19 days of work and no nights off!
It’s helped me keep my mind off all the other things which has been very helpful, really, but I am also looking forward to a night off!
Here is some more of my occasional series about constructing my garden railway: Grass Heart Vale.
Having considered ‘design concepts’ and tested out the general design (as covered in previous blog entries), the next thing I needed to contend with on my raised platform, ‘trestle’-style garden railway was how to make curves.
The YouTube video I was using as a reference for this type of railway (Richard from New Junction’s video here) indicated that cutting various slots/grooves in the trestle side frames would be enough to allow the frames to flex in a concertina-kind of way and seemed fairly simple. As I had quite a bit of this curving work to do, I decided that rather than attempt to saw loads of slots in all of the timber by hand, buying a specific tool for the job was a better shout and thus I purchased a Bosch Router for this reason. I’ve never used one before, but it is a very handy tool for such work, although a little scary on my first try!
Curve Experiment 1
For my first test piece which connected to my test section (discussed previously), I routed evenly spaced slots into both side frames with the gaps in the slots on the inner curve faces. I first tested routing at approx. 1/2 of the timber depth (10mm-ish of the 22mm), but that provided next to no flexibility in the timber to curve.
Curve Experiment 2
For my second attempt I routed the timber through about 2/3rds of its thickness (approx. 14mm of the 22mm). I think I also decreased the distance between each slot/groove so it still maintained some rigidity while allowing me to flex the timber where I needed to. This showed some promise and I then went about trying to set the curve to some OO gauge curve radius templates. To do this I used some temporary baton timber stakes to fix the curves at the radii I wanted them at by wedging the slotted timber frames with them. Although I could achieve a fair amount of curve and flex in the wood while the frames were in my arms, this approach turned out to be a total failure! As soon as I attempted to set the curve in the trestle frames parallel to my garden patio (a very gentle 90 degree curve), one of the frames would snap through the remaining bit of one of the slots and render the curve and whole side frame useless. This was quite a frustrating time in the garden railway build and I think took a few weeks to get right and quite a bit of broken wood!!
Understanding the radius of a curve
Lets quickly talk about the radius of curves, as I must be honest: I barely understood what any of this meant and the terminology of 1st, 2nd and 3rd radius curves, in model railway terms, confused me further. I worked out, at least, that the lower the number of the above, the tighter the curve was and I purchased various curve templates of the larger radii curves to see what would work best in the garden. It was these templates I was trying to match curves to when all the frames kept snapping.
After some research, I generally worked out that the radius of the curve is basically an orbit around a circle and working out the distance to the centre of this imaginary circle is what determines the radius of the curve. As I am building this layout in my garden, my limits of curves were relatively, well… limitless, so I researched the type of tolerances for curves I could expect from the locos I planned to run. This was primarily an issue for the larger scales I plan to play with on the single run of 16mm gauge track. Both O gauge and 16mm locos and associated rolling stock need to be able to get around these corners. 2.5 foot radius curves seem to be the preferred minimum radius for O gauge, which equates to 30 inches and in my simplistic head, meaning: the curve needs to orbit (at its tightest) around a circle that has a 30 inch measurement from its outer edge to its centre.
Next I had a great time running around the garden with a 30m long measuring tape and timber stakes to mark out what looked like weird clock faces on the planned curves on the railway. At this point in the build I had two main curves to contend with: the one at the front of the garden, near the house; and the one that will exit the shed, which has limited space due to its proximity to the fence.
I worked out that the shed curve could have a max curve of 32 inches on its outer edge and still line up (eventually) with the prototype section straight I’d already made. Also to note that it was at this point I worked out that both the outer and inner curves of the layout have different radii based on the width of my running platform top (I know, this is fun right? All the lessons in school I clearly burned from my memory or filed under ‘I will never use this in my life).
I’m still a bit mystified by those numbered radii curves that are often discussed in OO gauge model making, but I think they were designed for set track, which makes sense and they are:
1st radius = 15 inches
2nd radius = 18 inches
3rd radius = 21 inches
4th radius = 24 inches
It becomes a bit of a mystery from here on what number each is, but you can get ‘track setta’ templates for 30, 36, 42, 48 and 60 inch curves (and I ordered most of the wider ones for this reason).
Christ, this has all taken a lot of hard work and thinking (and we are no where near the end yet), anyway….
Unlike the restricted shed curve, the curve near the front of the garden could be an even wider curve/bigger circle, so I assumed maybe 40 inch curves was a good general rule of thumb for the rest of the layout and to keep things consistent (apart from the shed curve). Using 40 inch curves meant that almost every tolerance of a loco I would ever run is accounted for (and nowhere near their limits too). FYI, most OO trains are designed to run around somewhere between 2nd and 3rd radius curves, so there is no major problem there for running any OO stock on my garden railway.
The failures of Curving Experiment 2
So back to my failed attempt at getting this curve to work: The second version of the curve was tested with 6mm wide slots cut with a router bit at 5cm intervals and cut into the 22mm timber to about 14mm deep. This all aiming for that 40 inch curve which snapped every time I tried to do it, usually in the middle or near the end of the length. I did this a few times and got quite irritated. After all, in that YouTube video by New Junction it said this slot cutting approach was doable and sounded fairly simple to make a curve, so what was I doing wrong?
The answer was MATHS, I was doing maths wrong.
How to actually make curves: Kerfing
What I was actually trying to achieve with cutting slots in the wood is for the slots to close up on one end (basically the slots end up the shape of a pizza slice). The width of that slot determines the radius of the curve, while the spacing of the slots determines how gradual that curve is. To create a ‘pizza slice’ from each slot in the wood, you need to let go of any assumption you have to maintain the rigidity of the timber frames, you are slotting this bastard within millimetres of its life! Waaay beyond 2/3rds of the timber thickness.
It was around this point I discovered via internet research that the act of creating a concertina of wood via cutting slots and bending the wood into a curved shape actually has a woodworking term attached to it and it is called: ‘Kerfing’! That’s right folks, once you know this little factoid, this shit get’s waaaay simpler, because then you can find all manner of tutorials on this very topic! In order to curve your trestle railway frames, you need to understand the ancient woody art of ‘kerfing’!
It allows you to punch in details regarding the timber you are using, and the resultant curve you are after. These details include: the radius of the curve, the thickness of the wood, the width of the router slot and the angle around the curve you are aiming for and it will then calculate for you:
the length of timber required to complete the curve;
the minimum thickness within the slots (i.e. what you cut down to with your router); and
the distance required between each slot.
I know, right? A game changer!
And what can I say about this approach? Well, I used this kerfing calculator tool for every curve from here on in and it worked across the entire layout (and looks great)!
I opted to purchase a much smaller 3.5mm router bit to make the kerfing slots which provided much more gentle curves with very few instances of timber breaking.
A few important things to note here about kerfing timber for railway frames:
My frames are 22mm thick ‘rough cut timber’ and I’m leaving only 3mm(ish) in the routed slot, so you have to be SUPER careful when manoeuvring the timber around after you’ve ‘kerfed’ it. I mostly kerfed the timber with the router while the timber was rested on much larger, rigid pieces of wood to limit any damage once it became super flexy.
The rough cut nature of the timber tends to mean that only one wide face of the timber is ‘smoothly’ cut and this side absolutely has to be the one that has the 3mm of the ‘connecting’ bit of wood across the slots. if you use the rough side by mistake, you tend to find the slots just open up through the rough strips of the cut like a comb rendering it useless.
Some of the ‘rough cut timber’ is so rough I’d call it ‘ripped up timber’ and I have actually gone to alternative shops to buy it as some stockists haven’t got timber with a smooth side at all, rendering it near useless for kerfing. Wickes was particularly poor for this.
Included below are some screen shots of some of the curves I used from the kerfing calculator and they worked perfectly. My mind is still quite boggled as to how Richard in his New Junction YouTube video managed to just cut slots and achieve a curve after all my planning and prep. Madness (but also, well done)!!
The curve at the front of the garden ended up being 66 inch radius on its outer side and 56 inch on its inner radius. The difference between the two radii being the width of the railway running surface (234mm). For these reasons (and a couple of others) I concluded that the 16mm track would run along the outer edge of the layout to maintain a wide radius for larger locos.
Final set up of the curves
To finish off my tail: In order to ensure the wood was as rigid as possible I applied some strong bond external wood glue to each slot on the kerfed timber and then, very carefully, moved the newly grooved piece into its curved position. It’s a really satisfying operation, actually, just putting light pressure on the wood so each slot closes up and creates the lengh of curve that you were after from the beginning. I tended to do this curving with the bow of the curve facing down (thus the closed up slots facing up), which often meant, for example, you could ‘hang’ the curved wood between two chairs and allow all the slots to close up and the glue to set. Some times I put a weight, like a lump hammer, in the middle of the timber to ensure all the slots closed up. Once the glue had dried, I would often run another coat of glue through the closed slots to provide some extra fill and rigidity.
A video of a section being kerfed and the grooves closing up. Note that the work piece is sitting on a solid section, un-grooved bit of wood underneath to limit movement of the main work piece.
I then fixed the curves together using cross members as per the design on my test piece. This was quite easy as the space between the slots is often a small straight so, that is how I selected where to fit cross-members and meta post legs for the railway. To aid positioning in the garden, I pushed temporary stakes into the ground for the curved pieces to rest on while I set up the permanent posts.
The rest of the curves on the railways are mostly matched to a 42 inch inner curve and a 52 inch outer curve. Below are some shots of some of the final curves (both as a ‘skeletal’ under frame and the final versions).
Blimey, this blog has been a bit of an ordeal hasn’t it? I do hope you take some learnings from my extended dialogue on the mysterious art of kerfing AKA: how to create an attractive and accurate curve on your trestle garden railway.
Continuing my retrospective on the building of my garden railway and the proof of concept section I worked on last year, I next started work on building the running surface and final fascia around the ladder support structure. I used thick 18mm marine hardwood ply as the running surface for the layout as per New Junction’s video. Rather than try my hand at making perfectly straight cuts, I just got the 2.4m length boards cut to the correct width at my local B&Q to sit relatively flush over the top of my frame. This meant I got two 1.2m lengths and an offcut out of a single 1.22 x 610cm board. In hindsight, I think I would have saved a lot of money by cutting up larger boards, but meh… you live and learn.
I marked out the position of the bracing cross supports beneath the ply board top so I could easily screw the board down. I countersunk all the drill holes on the ply surface and drove(?) some wood screws home.
I have already received some comments on social media around using ply along the lines of “it will rot!!!!” And “it will decay and won’t last!!!” Indeed my friends, but please see my original principles: posted in my first build blog: it is a temporary (permanent) structure. It will indeed decay with time (as will I). I am okay with this. It will last the 3-7 years I require in any case, so I assume it will last as long as my shed will (the railway actually faired much better than my shed did over the very volatile wet winter of 2023).
The main issue with ply is how it warps when it starts soaking up water, I have attempted to account for this by sealing it with waterproof PVA, including all cut edges and drill holes. I then gave it a coating of fence paint to make it more water resistant (this all before it is covered in roofing felt), so I’ve probably done all I can, right?
At this point I also started painting the main frame and fence post supports a coat of my core colour with is a nice wood paint/wood stain with a satin finish, in a nice shade of maroon.
I then approached fitting some roofing felt to top it off. I picked some felt that had a good 1:76 scale gravel coverage (I spent a lot of time staring at roofing felt at close proximity in B&Q). I fixed it down with a nail gun on this prototype section, which didn’t work particularly well, and then switched to clout nails for the main build (as discussed below). I made sure there was some roofing felt overlapping the side frames too.
To finish the final look of the railway (and to instigate some safety features for miniature trains) I fitted some thin pine around the sides of the top surface, to act as a fascia or ‘lip’ to the railway top, which makes it look quite tidy and also prevents any trains from running off the top surface into the grass below. I only fitted it perhaps 0.5-1cm above the top surface as I didn’t want it to obscure photos of trains running (priorities). It roughly sits at the height of the 16mm gauge rails.
So there we have it, the first 2.4m section of Grass Heart Vale complete. I rushed outside with my lengths sporadic collection of Peco track to see how it will look with track on and also took my new Dapol Yeoman Class 58 out for a photo shoot on the section.
I was immensely excited, tiny trains in the garden!
This helped my visualise the scale of the task ahead of me and I could see some of the challenges to come. The section towards the shed would be complicated as it is where the garden is the most graded and there is limited space to curve the railway into the shed.
All in all, the design of the railway didn’t change significantly from this test section. I was nervous about the incident with the meta post spikes, as I don’t want to get stuck with a 4ft spike not driving into the ground as planned (and not being able to remove it after). I don’t think I had many instances of cutting the final height of the posts after they were driven into the ground again (as I did cut them a bit too low in places, which made final fitting of adjacent sections tricky).
I also eventually replaced the roofing felt top and redesigned how I fix it down on this prototype section. As summer progressed I saw the nails rusting on the surface and I also accidentally threw some wood-glue on the roofing felt while making other sections, so I ripped it up once the main loop was complete. Instead of the slightly crap job I did with the nails and the nail gun last time (seriously, my nail gun is crap, I do hope there are good ones around), I instead used ‘clout’ roofing nails with a wide head and exclusively fixed them down at the sides of the layout where they were eventually hidden beneath the fascia boards. This adds extra protection to the boards as there is now limited opportunity for water to soak through the roofing felt on the top surface.
So feeling good about myself I decided it was time to maybe do some messing about with curves, but that can wait until next time.
Yesterday I planned to complete the moisture barrier, but the sun appeared all of a suddern, and as anyone living in the UK in 2024 knows, this is something to be cherished and utilised as much as possible. We never know when we will see the glowing yellow ball again! Therefore, I shifted into my ‘sunny day back up plan’ and started work repairing the sides of the fascia of the shed roof. This panels were hit really hard by the rains of recent years and rotted away really fast. The front one broke apart last year and I managed to replace it in autumn. I knew the side panels also needed doing and I’d spotted a light leak into the shed over winter that appeared to be emminating from one.
I cut the wood to size (I used tongue and groove flooring as the replacement parts) and painted some wood rot treatment on the frame underneath and some rotten wood hardening varnish, which dried quite well in the sun.
I then added some bitumen adhesive tape over the work area to create a water proof seal and then fitted the replacement panels, which is always a challenge when working on your own, but I managed. I then added a final strip of bitumen tape to the top of the panel to prevent water from leaking between the roof and the back of the panel.
In moments of pause for paint drying, I turned my attention to the inside of the shed and the moisture barrier. The polythene was thicker than I expected and is actually much simpler to fit than I thought it would be. I just hung it up the wall and stapled it down and then cut away any excess causing me problems nearer the floor and then formed it around the walls, while I tried to remember where I’d fitted power sockets and vents beneath. The simplest way to release the various bits of equipment was to carefully cut an ‘x’ through the material in the shape of the socket/vent beneath and then push the polythene over it. Although it looks like I have created some murder cabin at the moment, I’m very pleased with the speedy progress of all these jobs and it is quite nice to be in the shed and not feel all the prickly rock wool.
I’ve also worked out a plan for the flooring, which I’d overthought a lot already, but it seems that PVC floor tiles for workshops might be the way to go and sounds like quite a quick fix.
After a very long break from working on my Class 13, I decided to return to finish off the job over the winter. I feel that me from the past must have been aware that finishing this deceivingly simple-looking project would take quite a while. I was a little surprised it took most of the winter months and into the spring to finish it, but it did. I also took the opportunity to complete and finesse various projects that needed some work in the same colour scheme.
Picking up where I left off with the BR Class 13.
First off was deciding on a method of connecting the main multiple working cable connections between the master and the slave units so that they wouldn’t weaken or break when running the two units together. The solution was something that was likely unavailable to me at the time I started working on this project in 2010 (I know, shhhh!!) and they were tiny, tiny magnets! I basically glued some tiny disk magnets on the face of the multiple working connectors and made up a bit of flexible wire and soldered it to the sockets. This allows the locomotives to move and the wire to take tension but not put too much strain on the cable.
The multiple working connectors made from magnets and flexible wire to help movement around layouts
I also added some handrails and details around the cut-down cab of the slave unit and added some extra layers of flooring as I didn’t like the gap on the running board. I then started masking off the warning stripes and windows so I could do a full respray of the BR Blue and renumber the locos. I used various masking tapes and masking fluid. I primed the parts with a mix of Halfords grey primer (for the plastic parts) and etch primer (for the metal buffer beams and the rear of the cut-down cab). I then did some pre-shading with Vallejo black around various panel lines, not sure whether it would make that much difference.
I then plucked up some courage to airbush the body with LifeColor BR Blue, it took many layers of blue, but, to my surprise, the pre-shading did make a difference and gave some depth to the bodywork, which really looks superior when it is next to the flatter colour of RTR painted stock. I then gave the models a gloss varnish finish.
Airbrushing progress
When that was all cured, I did some more masking and concentrated on the big yellow bufferbeams, and friends, this took ALOT of time! Mostly because I started with the wrong shade of yellow. I first used a railmatch acrylic shade that was actually the apparent correct early shade of warning panel yellow, but the yellow was far too cold and made the Class 13 look like I’d painted it in National Coal Board livery. I then tried a later shade of warning panel yellow that was a bit more yokey, and although it looked better, it didn’t match enough with the warning panels I’d masked off earlier (I was trying to avoid having to repaint the factory-painted wast stripes). After doing some tests with various yellows, I found that the closest match was another Rail Match yellow, this being a special ‘faded’ version of warning yellow, but to make matters more complicated, it was only available in enamel, which meant it would likely destroy all the layers of acrylic-based yellow I’d already applied to the buffer beams (scream on my behalf please)!
Various shades of yellow, mostly wrong and finally getting to the right one.
I took a long deep breath and re-masked the loco and re-primed the buffer beams etch grey and painted the enamel yellow onto the bufferbeams. It ended up much thicker as a finished coat than I would have preferred, but it was that or stripping the paint off. I also added BR double arrow emblems and a new number to the Class 13. I made it a transition numbering of a fantasy 4th loco as ‘4503’ mostly because the donor bodies of 08s I have used couldn’t easily pretend to be any of the original three. My numbers were a mix of water slide transfers from Fox transfers and pressfix ones from HMRS. Having read the instructions for the HMRS ones, I didn’t realise that you could actually remove the transfers from the main backing card first!!! This means you can actually see the transfers when applying them!!! Having done this ‘blind’ previously, I must admit this time it was a lot less challenging! Do I get some advanced, yet slightly foolish modeller prize?
Fitting transfers
Next, I picked out the handrails with white and painted the extra large buffers, made up some vacuum pipes from guitar string, and tidied up some areas of the paintwork. I also fitted some brass coupling bars to connect with NEM couplings.
Finally, I decided to revise the wiring between the locos, as I’d previously installed a removable socket, but decided this was a bit of a waste of time, and decided on some 36 AWG gauge wire (I don’t know what that code means) in black and fed it through between the slave and master units and connected them to my revised pick-up board (which I assume I discussed in a previous installment of this blog, I’ll check).
The connections between the units and my revised pick up system.
Then it was time for some test running on the carpet railway and after some minor pick-up adjustments it ran great!!
Test running on the carpet railway
So there we have it, another one of my model projects complete!
The finished loco
Here are some links to previous installments in the Class 13 build.
A while ago I threatened to update my shed to be a more enjoyable environment to work in over the wetter months and over winter. I talked about it here and discussed some inspirations for the upgrades. At the time I intended to convert the shed into a paintshop with a permanent airbrush setup. I think I also considered having it as a key location for my 3D printers too. I then outlined a bit of a ‘modelling manifesto‘ in early 2023 and the shed has now become the control hub for my garden railway project. Nonetheless, most of these concept usage cases haven’t stopped me from progressing with the enhancements.
How the shed looked quite some time ago now
The first job I tackled a while ago was ripping out the potting shed shelf, as I knew I’d want to have different work surfaces installed. I then filled any gaps in the shed with expanding foam. I then started laying out wiring for power sockets and lighting including a small consumer unit for sheds. I had a great time doing this. most of the sockets are supported on pieces of wood that lift them away from the panel surface of the shed. I then started to fit ‘rock wool’ between the shed panels. Sometimes holding it up with gaffer or aluminium tape so it didn’t fall straight back out. Cutting the rock wool into block shapes is quite easy with a plasterboard saw. cutting large slabs of rock wool reminded me of sheering sheep, which was quite amusing. The most challenging part of the insulation job was doing the ceiling of the shed. I tried holding the rock wool up with tape, but it often fell down, so I opted for nailing thin strips of batton over the top of the rock wool to keep it up on the ceiling. That isn’t always an easy job while working on your own; quite a few times the wood fell down and smacked me on the nose!
The shed ceiling with batton supportsInsulation and wiring fitted.
I also fitted two sets of outside sockets to the front of the shed and added them to the ring main, so I had some options for powering lawnmowers and things. I had also bought all of the materials to run armoured cable up the main house consumer unit, but decided I really didn’t want to be responsible for this step, so decided to ask an electrician to do this and then do a final test of my shed wiring. After a few quotes and a surprisingly long wait to get the right electrician, he happily sent a day laying out the armoured cable to the shed. Apparently, my wiring was mostly okay, but he suggested I should have left longer lengths of cable in each of the sockets, so that is a good lesson learned for the future. There was much celebration when the time came to switch on the lights. It is quite an exciting thing to have power in the shed after so long!
Shed power!
Now with the power in the shed, it was time to finish the remaining jobs and complete the insulation project. I installed x4 circular vents into the shed, x2 on the front, and x2 on the back, and worked out how long the piping would need to be to work its way through the insulation layer and the final wooden panel surface. these vents are also great as I can fit an expandable pipe from my airbrush extraction unit into them. I then stuffed the remaining rock wool around the panels I hadn’t completed including around the consumer box.
The vents installed.
The final panels in the shed that needed insulation were those where the garden railway enters and exits the shed on the left and right shed walls.
I fitted the superstructure of the railway through the shed last year, to allow me to complete the main loop, but I am still yet to open up the holes in the walls for the tunnel entranceways. At the moment there is just a slot on either side for the inner frame. I’m quite impressed with myself that I have held back from opening up the tunnels, but there were practical reasons: rain and small animals getting in the shed being the main reasons.
For insulation to fit around the tunnel mouths though, I now needed to at least setup an inner frame for them. I had previously worked up some calculations for the maximum height of railway stock likely to go through the shed and it seem like 5 inches for OO and 7 inches for 16mm seemed safe enough, so I set the height of the railway surface at 7.5 inches and created a batton frame overlapping the main railway frame as much as possible. I had to cut some slots out of the wood at certain points to allow the railway to fit. This was more challenging on the second exit point as it is on the tightest curve on the railway (about 32 inches, so not that tight compared to most curves). Also, don’t ask me why I’m working in inches today, I don’t really know why.
I blocked the under neath of each frame with a large piece of wood to minimise any gaps leading into the shed. I plan to use some steel wool to fill any smaller gaps and probably some sealent from the outside to finish it off.
Shed entry way shots
I then started pressing the remaining rock wool around the frame and between the shed panels. Although quick to write, this all took quite a while, particularly as I’m still recovering from my second round of covid.
I did manage to sit in the shed for a bit and drink a coffee and finally feel like the space was starting to come together really well. I can’t wait to see trains running through the shed, it’s going to be ace and perhaps that isn’t that far away now!
The next job is to install a moisture membrane over the insulation.
Grasslands Models 