Front Garden Railway Rob

This page last updated: 6 January 2024

In May 2014 I decided that I needed a front garden railway.  I don't have much of a front garden, at most 3.5 metres by 5.5 metres and with a 0.5 metre slope across the short dimension.  But I didn't have anywhere else to put a railway and so it needed to be done.  The space is so small that realistic curves are impossible however I did want to run real steam so I commissioned Roger Melton of Just The Ticket railway supplies in Salisbury to build me a Gauge One (i.e. 45 mm track spacing, with an accepted scale of 1 metre == 32 mm) loco that would fit.  And, of course, it had to be a 0-6-2T tank engine of the type that would have run on the Rhymney Valley line in South Wales in 1964, the year I was born.  In discussion with the secretary of the Welsh Railways Research Circle I determined that the "native" 0-6-2T engines of the Rhymney Valley line, with the flat-topped tanks, as seen in the Rhymney Valley line books, were all withdrawn by the mid 1950s so the engines that would have been running on that line in 1964 would be ex-GWR 0-6-2T 56xx series engines, in BR colours, with slanted-top tanks.

Roger said he couldn't commit to a timescale, so the engine has been in the making for a little while.  Provided it will go around the tight radii I don't mind.  Now that he's not far off completion it is time for me to begin work.  Below find my front garden railway diary, starting with the things I did in the years of waiting and then the clearing of the space. Updates on progress are then in chronological order, most recent at the top.

As an interesting example of synchronicity, after this project started our local pub, the Railway Arms, which had been closed by Charles Wells (who were planning to turn it into houses) was bought by the community and is now run as a community pub, in which we are of course shareholders.  And the logo chosen for the Railway Arms includes a 0-6-2T tank engine:

Railway Arms logo
[logo reproduced with permission]

Track
 6 January 2024

Despite having already purchased a load of Peco G45 track, after meeting Cliff Barker back in 2017, I have now purchased his stuff for rather better scale accuracy.  Comparing the two, the differences really are huge:

Peco v Barker
Peco v Barker
Peco v Barker

They almost look like they must be different gauges, yet the spacing between the rails of both is 45 mm.  The truth is they probably are different scales, the Peco G45 track intended for rather larger scale but narrow-gauge rolling stock.

The practical difference is the height from the tops of the "chairs", i.e. the things holding the rail to the sleepers, and the top surface of the rail, which in the case of Barker is 2 mm and in the case of Peco is a truly gi-normous 4.5 mm.  In other words Peco track could run just about anything.  I have checked and all of the wheels I have run perfectly happily on the Barker track.  Specifically, a chose his "fine" version, in stainless steel, with a very slightly widened track spacing of 45.5 mm; that will help the running on my unrealistic radius.

As shown in the plan below I needed two points: one straight right-hand and one curved left-hand, both at my unrealistic radius.  I considered making my own but the process is quite involved and includes soldering rails in a jig that I'd need to make, etc., so I gave in to sense and got Cliff to make them for me, adopting 1121 mm as the radius of the outer rail.

The
                made-to-measure points
The curved point in position
Curved point in perspective

Since he only lives 50 miles away from me I dropped into his "shop" (i.e. his garage) where he was busy with one of four injection moulding machines pumping out track segments while a gentleman at a bench made points for other people; fascinating.

To lay the curves of track I needed a radius guide, 45 mm wide, also of the exact 1121 mm radius.  I created this in VCarve, the design SW that happens to come with my High-Z/S-400T CNC milling machine (from cnc-step.de); I wanted the guide to be 750 mm wide but the machine is limited to a 400 mm throw so the design had to be "tiled" (which VCarve does for you) and milled in three separate sections, moving the 2 mm thick aluminium sheet exactly the right amount each time (by dint of having marked and drilled the temporary fixing holes very carefully).  I used a Mastermill AL (HPC) 3 mm slot-mill from Europa Tool Co Ltd, run at 14000 RPM, applying LOTS (slightly over one 300 ml can) of WD40 during the cut to stop the tool just melting the aluminium.  A sacrificial strip of 3 mm ply was placed underneath the job so that the tool could cut through.  In the pictures below you can see the test cuts and mistakes I made through my own incompetence and while having a few arguments with the collet-holding mechanism of the milling machine, swiftly resolved with the excellent customer service of Philip at Prototools.

Radius guide design
Milling set up
Tile 1 completed
Tiles 1 and 2 completed
WD40 in action while tile
                3 is being cut
All
        three tiles milled
Completed radius guide

If you look carefully you can see very slight stepping in the radius guide at the edges of the milling "tiles":

Stepping due to tiling

Measured, the step is somewhere around 0.25 mm, which should be OK I think, especially as, through sheer luck, the steps are symmetrical about the centre of the guide.  Time to start thinking about concrete.


Mountain, Oh Mountain
24 September 2023, updated 1 October 2023

As the old Vogon poem "Mountain, Oh Mountain" goes, I needed to decide where to put the mountain of the front garden railway. Since I want to have a tunnel, there has to be a mountain, but it should look as little as possible like I just needed a tunnel and so piled some earth somewhere.  Yes, I know, anyway...

To better visualise the problem, I spent the afternoon laying out where the track might be using 3 mm ply and spare decking board; ignore where the earth is currently piled, that's just where it ended up when I dug the hole for the water tank.

Of course, there also needs to be a station, or maybe just a "halt" given the lack of straight edges; enough for the two coaches of DMU I have.  I had always assumed this would go at the back, on the concrete surrounding the bay window; need to ponder.

Rough layout, front view
Rough layout, rear view

Update, 1 October 2023:

I really wanted to have two sidings, each with a clear 2 metre run to them, so I rearranged somewhat, bringing the left-hand curve as close to the yew tree as possible to allow it.

With other siding

 

Loading Gauge
10 September 2023

Now that the engine is confirmed to be able to take the curves it is time to start thinking about construction; first, given the unrealistic radii, I calculated the loading gauge.  I had built the Gauge One DMU some time ago for just this purpose; taking its measurements and reckoning on a 2242 mm diameter, we have:

Loading gauge

Result: I need to allow an extra 22 mm clearance on the inside of my curves and, for safety, lets say half that on the outside of my curves.

Can She Take The Curves?
27 August 2023

Having shown that the engine runs, the next question is whether she can take the extremes of curve radius required.  I brought out the circle of Peco G45 standard curves, with a 0.6 m diameter radius, a truly unrealistic curve (about half what the front garden railway will have) that I used to test the chassis of the DMU, with similar results:
 
Conclusion: yes, just need to unstick the engine startup (and remember to mount those pointless rear wheels in future).

Running On Compressed Air
26 August 2023

After a few exchanges of instructions and parts with Roger I now had the bits that fit into the hole in the cylinder block, exhibits A being the "ARM1G block lubricator banjo bolt", "lubricator banjo" with pipe attached, and a brass union to be soft-soldered onto the end of the pipe, steam/oil flow path indicated in the pictures by the arrows:

Lubricator Banjo Bolt
Lubricator banjo
Assembly

...exhibit B being a lubricator reservoir:

Lubricator reservoir

...and exhibit C being a double-union that allows one to be connected to t'other.  I took a few guesses as to what connected to what, then called Roger to find out how it really works: the pipe from the cylinder to the reservoir is a kind of bi-directional thing, steam being pushed out and oil dragged back in from the reservoir; the second, blocked-off, pipe from the reservoir is purely for draining.

The pipes had to be bent so that the reservoir would fit inside the engine body, where the water tank would be in a real loco; Roger's advice was to do the bending by hand but to "do it once", as the copper case-hardens quite swiftly.  After doing the bending, soldering the union onto the end of the pipe was pretty straightforward: the rear-end of the union is designed to fit tightly onto the copper pipe and, with the surfaces well cleaned and fluxed, applying gentle heat with a blow-torch, soldering took but a moment.  Here are the parts, bent, soldered and ready for fitting:

Lubricator assembly bent and soldered
Lubricator reservoir, with double-union, bent and
                ready for fitting

...and here they are fitted:

Lubricator
                assembly fitted, bottom
Lubricato
                assembly fitted, top
Lubricator
                reservoir fitted

With that I brought the stage zero equipment into play for a static fire and, hey presto:
 

Water Update
12 August 2023, updated again 26 August 2023

I took a look at the water level in the tank today: with the water intake still blocked for the entire life of the water-flow experiment so far, my dip-stick showed 15 cm of water.

Dip
            stick

That means there are around 185 litres of water still in the tank, which must all be rainfall collected via the return from the pond at the far end, and this despite the pump running 12 hours a day for 3 months over the summer, with a slight leak from one of the joints.  Since there is plainly no shortage of water, I upgraded the pump to 1000 l/hr, made a proper job of the joints to remove the leaks, then unblocked the intake: time to see how things work for real.

One thing I have noticed is that the water-return from the pond needs upgrading: a leaf can block the small pipe entrance; it needs to have a filter on it so that detritus doesn't get in and, at the same time, have a wider filter-mouth so that collected detritus on the filter surface does not itself cause a blockage.  The same goes for the overflow, which will be the next thing to test.

Update 26 August 2023: water seems to overflow correctly:

Overflow
        working


Ystrad Mynach Station Plans Found
19 July 2023

In the 70's, when I were still a boy, and wanting to make an N scale model of the local Rhymney Valley line Ystrad Mynach train station, my father sought-out the ex-station master, Mr G Morgan, and got him to draw a plan of the station.  I duly made some plasticard models of the buildings, which have survived and can be used as a basis for sizing, since I would have measured the buildings at the time (all of them are long gone).  Last night I found all of this hiding in a box in a filing cabinet.

Envelope
Building
                sketches
Track sketch
Main building
Main building,
                toilet block
South platform
                shelter
Merthyr branch
                platform buildings
Footbridge

Preparing To Run On Compressed Air
28 May 2023

It took a little while to figure out the thread that was on the inlet to the pistons at the front of the engine:

Thread parameters

Such a large number of threads per inch could not be found in the usual tables; turns out that this is a 1/4" model engineers thread; you probably need to scale your threads to carry steam pressure.  It took a little while to purchase the right size of union cone, nut and copper pipe:

Brass parts Pipe
Union cone and
                nut

I cut a 50 mm length of the 1/4" copper pipe (20 gm copper pipe, meant walls about 1 mm thick) and drilled it out, 3 mm deep, to 5 mm diameter in order to fit on the rear-end of the union cone.  I fluxed both the drilled-out section inside the pipe and the rear-end of the union cone then, with a blow torch, tinned the fluxed-inside of the pipe and, while balancing the cone/nut on something supportive/flame-proof, joined the two together by applying blow-torch.

Drilling out
Solder in pipe
Fluxed union
                cone on nut
Soldered
Soldering
                station
Adapter completed

Blowing into the pipe proved it was air-tight, however it also showed that there was a part missing on the other side of the cylinder block which wasn't in the box: Roger had said this might happen and to contact him when it did; more once I've got hold of that.

Missing bit


Water Flow
21 May 2023 - 28 May 2023

While the parts I needed to run the engine on compressed air were being sorted, I installed the pump etc. for the river, so that I could check that water goes from and to the right places, viz:

water
          flow

...and also to get some idea of the loss to evaporation etc.
 
Water flow
                experiment, left Water flow
                experiment, right

[Ignore the green conduit: it contains the broadband cable].

A dipstick showed that the water in the tank was 350 mm deep at the outset, all of which was collected off the roof in just a few hours of heavy rain.  The tank is 1010 mm x 1250 mm with walls probably 15 mm thick which suggested 0.43 cubic metres of water or 430 litres; 1.23 litres per mm.  With the system running in steady state, the little pond full, this dropped to 320 mm, meaning that the pond etc. held 37 litres.  I set it to run from 08:00 to 20:00.

Update: after seven days of running for twelve hours a day, in reasonably warm weather, and with a drip from the connector out of the top of the tank to the "river" pipe [which should serve to make this a worst case measurement], the water level in the tank was 305 mm: 18.5 litres lost, about 250 ml per hour or, assuming all of the water in the tank could be used, the system would run for about 4 months from one tank-full.  Given the apparent fill rate it doesn't look like there will be any problem with water supply.

Training Day!
15 April 2023

At last, almost 10 years after I began, training day: I picked up the bits of the 0-6-2T tank engine from Roger Melton at the Peterborough Garden Rail show (the annual show of the 16 mm society).  There is slightly more bodywork assembly required than I had anticipated, so I didn't hand the lot over to Fosworks after all.  The main point is that the boiler has been tested, the motive mechanical parts all work and, critically, there is no flange on the centre wheel, which should allow me to run on the unrealistic radii I have to use.

The main missing part is the mechanism through which the boiler is fired, the gas vessel.  But first things first: I need to find out the size of the fitting at the front of the engine, the one feeding the cylinders, then I can drive it from a compressor to demonstrate that (a) it works and (b) it travels around the required radius.

The bits
Wheels
More
                assembledish
Fitting Cone and nut on connection from boiler

Nearly But Not Quite
12 March 2023

Today was meant to be training day, when Roger was going to hand the engine to me at the Warwick Garden Rail show.  However, he had thought we'd agreed the handover to be at the Peterborough Garden Rail show.  Ah well, one more month (15th April) doesn't matter and, in fact, Fosworks, who I will ask to fit the remote control, will be at the Peterborough, show so I can hopefully pass the engine from Roger to them, saving me a trip to Lancashire.

As a consolation prize, I spotted that Malcs Models had a Mamod Brunel vertical boiler engine which runs on Gauge One track, so I bought one of those and spent the afternoon fitting and staining the boiler cladding in a somewhat rustic fashion.  This is also something to practice my gas-loading technique on.

Mamod Brunel
Mamod Brunel

It is a really good example of the most simple steam engine:

Mamod Brunel labelled

On the right of the engine is a gas vessel with a small inlet valve built into its top surface.

Mamod Brunel gas vessel

Mamod supply an adapter to a fit a perfectly normal butane gas canister, as one might use when camping, which one shoves into this valve to fill the small vessel with gas.  The outlet valve can be opened by hand to let the gas flow into...

Mamod Brunel gas burner Mamod Brunel as burner
                lighting slot

...a gas burner underneath the engine, just like you might find on a gas hob, which can be lit through a slot opened in the side of the chassis.  The gas heats the water in the vessel that is in the middle of the engine, the boiler, which you will already have filled with water:

Mamod Brunel water vessel,
                right
Mamod Brunel water vessel
                left

The outlet valve can be opened to release the water, as steam, into a piston which drives the wheels.

Mamod Brunel piston

So a bit like a kettle on a gas hob with a pipe going from the spout to a piston.

Less Tree
12 February 2023

Things are about to get exciting: Roger hasn't been able to get the engine painted (even after the fourth engine-finishing guy took over) but we've decided that he'll just deliver it to me as-is and I'll do the rest; I will pick it up on 12 March.  Time to get arse into gear so, as a demonstration of arse'dness, today the yew tree was halved in size.

Before After

Lets hope it survives the amputation.


I Made A Hole
9 January 2022

2021 wasn't engine year after all (Roger is now quite positive about 2022, having found his third engine-finishing guy) but I didn't want to waste my exercising and so I made inroads into the river system [assuming you can make a road into a river]: there needs to be a river/brook/dribble running down the middle of the layout, I had decided, and hence I ordered a 500 litre square/flat water tank from Ecosure, designed to be buried (I have an identical one buried in the back garden) and, on the hottest day of the year (13th June), I dug most of the hole in the front garden to put it in and then followed up on the August bank holiday weekend with the remaining depth (until I reached the chalk) to fit the tank in nicely.

First excavation
Finished hole, down to the chalk layer
Tank in 'ole
'oled

I bought one of those automatically-levelling laser levels, which came with an L bracket onto which it was a [tripod] screw-fit.  The L bracket had magnets on the back so I could drive a length of angle-iron I happened to have lying around into the ground as a mount.  I set the level to be just below the paving at the rear of the garden area.

Laser level
Laser level on post
Level at rear of garden Level at front of garden

I purchased a plastic washing up bowl, which included a handy drain-hole in the bottom, and cut the sides of the bowl so that I could slide it into position on top of the drain under the down-pipes visible in the rear of the picture above.  Then I 3D printed [in ASA for ultra-violet hardness] a part to sit underneath the drain with a nozzle to lead the captured water into a length of 19 mm inside-diameter hose.

Rain-water capture
                washing-up bowl
3D printed drain
                part in blender
3D printed drain
                part in ASA

I purchased two float switches, a 12 V latching changeover relay with mains-rated 50 Amp contacts, a small 240 V/12 V transformer, a bridge rectifier, some 100 uF capacitors/10 kOhm resistors, one mains plus two 12 V indicator lamps and two momentary action push-button switches to make myself a circuit that would switch on and off a 450 l/hour pond pump (the kind designed for small water features) [later in the year this was updated to a 1000 l/hour pump] depending on the water level in the tank.  The mains supply to the contacts of the latching relay, and hence the pond pump, was from a mains timer unit so that it could be set to come on only during the day or on weekends etc.  I bought a box to put it all in and spent the time between Christmas and new year putting it together.

I thought that it would be best to sort out the river bed as a free-standing thing, so that I could test it and ensure good/leak-free circulation before surrounding it with earth; to make this I bought myself a MIG welder and, as welding practice, I bought some lengths of stainless steel to construct a frame, 500 mm high, that could be lowered into the neck of the tank to hold the float switches and the pump.  The float switches were held on daisy-chained cable ties so that I could adjust their height.

Schematic
Circuit
                board, front
Circuit
                board, rear
Boxed up Floats and
                pump in frame
In the
                hole

Next: the river bed itself.

Railway Exercises
4 December 2020

You may have noticed that there has not been a great deal of progress. I didn't want to begin construction until I had the engine, 'cos I wanted to be sure I would build something it can run on.  In the intervening time I've purchased a 3D printer, designed and constructed a Class 116 DMU (since serialised in the Gauge One journal) intended to act as a loading gauge, which was quite the project in itself.  But I still wanted the engine before I started digging.

Roger tells me (told me a few years ago actually) that the engine chassis is done, and has even been run at a show, but it has not yet been finished (painted, lined, etc.); finding someone to do that job properly is what is taking the time.

Anyway, he believes it should be done next year so it is time to prepare the ground. And by that I mean me: I have to be in the right shape for construction, so Alice has helped me set out the exercise regime below for the winter.  I will complete the columns to show progress (if any) over time.





Actual Number In Set or Duration
Exercise
Description
Number of sets
Target number in set
7/12
9/12
11/12
14/12
16/12
20/12
21/12
23/12
27/12
28/12
30/12
3/1
5/1
7/1
9/1
11/1
14/1
16/1
18/1
20/1
23/1
27/1
28/1
31/1
3/2
4/2
5/2
8/2
13/2
14/2
22/2
24/2
Squats
Stand with your legs shoulder-width apart then, while keeping your heels on the ground, bend your legs as if you were going to sit down, onto your ankles; keep your back straight (not upright) and stick bum/arms out for balance.
3
6 to 15
10
10
10
12
12
12
14
14
14
15
15
14
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
Leg Lifts
Stand straight then slip one leg behind you with toes on the ground; raise that leg, straight, as high as you can behind you while keeping your hips and the other leg straight, before returning your toes to the ground.
3
(with each leg)
6 to 15
10
10
10
12
12
12
14
14
14
15
15
14
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
Press-ups
Obvious.
3
6 to 15
6
6
6
8
8
8
10
10
10
12
12
12
14
14
14
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
Rowing/pulling
Lie underneath something that you can use to pull yourself up (e.g. a broom handle between two chairs); with your arms at waist height, pull your body weight up.
3
6 to 15
6
6
6
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
The prank
Rob's version of a plank: a press-up but held at "half mast"; hold for at least 20 seconds, increasing up to 90 seconds.
3
1
20s
20s
20s
30s
30s
30s
40s
40s
40s
50s
50s
50s
60s
60s
60s
70s
70s
70s
80s
80s
80s
90s
90s
90s
90s
90s
90s

90s
90s
90s
90s

Other guidance:
Let's hope this makes a difference.

20 Ton BR Brake Van
6 January 2018

Over Christmas and the new year I have completed the Northern Finescale BR 20 ton brake van kit.  I'd guess it took about 15 to 20 evenings to complete.  I made the roof removable with a few magnets to hold it in place so that I could mount the battery box for the rear LED light inside (affixed with velcro pads), running the wires underneath the model.  The weathering part was much more pleasurable than I expected: just spend your time idly dabbing with this and that colour until the model looks "used".  Since I've painted the vehicle bauxite and labelled it XP I also bought and attached vacuum brakes so as not to be drummed out of the G1MRA scale accuracy club.

Brake van kit completed
Removable roof

The model was painted as follows:


Ready For The Winter
17 December 2017

The front garden is now covered in weed fabric and lying fallow for the winter.  However, I have made a small amount of progress in that I have constructed and painted the track-side hut kit.  I cut out the window panes with a small Dremel jig saw, used the kind of resin you put on glass fibre to glue it all together and then painted it as follows:
Covered in weed
                fabric
Trackside hut
                painted, human's eye view
Trackside hut
                painted, bird's eye view

Gauge One AGM
14 October 2017

Since it was being held just up the road from me in Huntingdon, I decided to visit the Gauge One society AGM; not to go to the AGM itself, of course, but to meet useful people in the show on the same site.  Here are the useful things I did:
Water towerInside water towerTrackside hut

The water tower will allow me to actually refill my engine from it (fitting a plastic hose between the filler pipe and the up-pipe and running a hose from the up-pipe at the base to a supply tank); the trackside hut will be a nice simple first kit to build.
Brake van kit
16
            ton grey steel wagon
While talking with Cliff Barker about the track I asked about remote control of points and he talked of a design they had in mind but which was too time consuming to produce.  It sounded really cool and so he pointed me at the Model Electronic Railway Group.  Celebrating its 50th anniversary this year, MERG is a UK-based model engineering group focussed on the electronics and computing side of railway modelling. How excellent - I joined immediately.  What a productive day.

Cleared
24 September 2017

After a few hours each over a few weekends spent with pick-axe and fork, and as many trips to the dump, the front garden is cleared. The large stones were saved to one side and five frogs, one toad, one bumble bee and a newt were rehoused.  Fortunately there is good soil right down to pavement level, which gives me lots of room to get imaginative with a river I have in mind.  Now I will leave it over winter, covered to avoid any unwelcome growth.

Cleared

Clearing
27 August 2017

As I've been waiting for the engine for a little while, the front garden has become somewhat overgrown.  So I've begun by clearing it; excuse the statement of political allegiance, I took the "before" photo in the run-up to a general election.

Before clearing
After clearing

One thing is evident: I will need a cutting at the back and a viaduct of some form at the front if I'm going to have a dead level playing field.  As you can see, I've trimmed the yew tree to give me room to run underneath it.

Slope under
          tree

Things I Did While Waiting
27 August 2017

While I've been waiting for the engine I began a plan using WinRail X2.  This is just a rough idea, laid out using Peco G45 track (which I have also bought), and doesn't take into account the position of the yew tree.  The inner oval is not part of the implementation, it is simply a standard oval of Peco G45 track for visual reference (since I can set such an oval up in the loft).

Plan, version 1

Also, while visiting a model railway show in Bishops Stortford, I came across a chap who made station signs; I had one made (we live on Victoria Avenue):

Station sign


Back to Meades Family Homepage