Rewanui - a 1940 West Coast NZ layout in 1:64

[Lawrence@NZFinescale]

I've had my arm twisted to display layout progress.  This presupposes there is progress to display, so the pressure is on a bit.

One of the many hitherto largely unaddressed challenges has been the track - or specifically the turnouts.  Nobody makes Sn3.5 track (16.5mm gauge) that looks anything like the real thing found locally.  As most of my turnouts are in a yard and mostly well covered in ballast it probably doesn't matter all that much, but the switch area is always empty of ballast and so I wanted to get that area right at least. As usual this resulted in going somewhat further than entirely necessary.

It will always be a compromise.  For historical reasons I'm stuck with a somewhat narrower gauge with wider flangeways than true scale.  As a result simply scaling down the real thing doesn't work quite right.  I have good drawings for 1926 70lb turnouts so they are the basis, but there's been a bit of adjustment to capture the look without necessarily getting all the dimensions dead right. I'm using code 82 rail, which is 2.08mm compared to 1.93mm true scale. Code 75 would be almost dead on, but I cannot get that in steel, which is what I intend to use for the final track.

I've made all of the turnouts before with copperclad sleepers.  OK, but I've moved on a bit and they don't really satisfy.  So I thought I'd try 3d prints.  The result is really good.  This is the prototype so there are one or two minor things to revise, but not a lot.  I was concerned about longevity and stability, so brass PCB rivets have been fitted and the rail soldered to them from below.  It's rock solid with no visible solder.  I fill the rivets with solder, flux the interface with the rail and then use the resistance soldering machine to make the joints.  The rivets are just the right length and the resistance solderer only makes contact with gentle pressure.

I deliberately printed the guard rails and bits of the frog as a) these areas get polished when cleaning track but they shouldn't be, and b) it enables detailing to be easily added.  The guard rails are slightly lower than the running rails so will not get cleaned.  The frog is flush at present, but one of the things to look at.  I've also captured the wide variety of rail fixings used. They are very fine and a bit vulnerable during construction, but once finished should not be subject to stress.

One of the keys here was printing rail filing jigs derived from the turnout model.  All filed rail was dead right first time more or less, though I did give one or two parts a second lick.

I'm undecided whether to simulate woodgrain in the prints.  It will be mostly wasted effort, but we'll see.  There's still some thought to go into the headrods (tiebars) as well, but many others have sorted that problem so it should be straightforward.

A fair bit or work to derive the models (1:7.5 and the Y shown here). However, very quick to print and fairly rapid and relaxed to build.  I only need 6 turnouts, so it should be quick once the last few issues are sorted out.

[finescalerr]

That may be the most impressive non-commercial track I've ever seen. Adequate. -- Russ

[Barney]

Nice fine bit of rail - just as it should be - real quality
Barney

[Ray Dunakin]

Amazing!

[Bill Gill]

Excellent!
Too bad there isn't any commercial code 75 steel rail lie you want.
Will all 6 turnouts have the same frog number or will you construct each to fit a specific arrangement?

[Stuart]

Fine scale modeling to say the least!

[Lawrence@NZFinescale]

Excellent!
Too bad there isn't any commercial code 75 steel rail lie you want.
Will all 6 turnouts have the same frog number or will you construct each to fit a specific arrangement?

Slightly frustrating on the rail as there used to be.  But one moves on.
The NZR was not big on custom track structures.  There was probably no need as the railway came first and so did not need to fit into tight spaces unless constrained by geography. Rewanui is constrained but all the turnouts are 1:7.5 other than the Y at the yard entrance.  Should be 12' swiches I think, but I've modelled the 10' version as there is a little bit of compression in the model.  So other than the Y they are all the same - other than being L/R which is trivial with printed parts.

[Bernhard]

Perfect work, Lawrence. I must remember the idea with the rivets!

Bernhard

[Lawrence@NZFinescale]

Perfect work, Lawrence. I must remember the idea with the rivets!

Bernhard

I stumbled on them really.  Just a couple of dollars per thousand (including shipping) from our Chinese friends on AliExpress.

[Hydrostat]

Lawrence,

really impressive work. I'm a bit lost about the rivets: do you solder them to the rails nefore you add the sleepers or with the sleepers in place? The latter is hard to imagine because heat would deform the printed parts, wouldn't it? And first way means to work very exactly, maybe with some kind of a gauge?

Cheers,
Volker

[Lawrence@NZFinescale]

Lawrence,

really impressive work. I'm a bit lost about the rivets: do you solder them to the rails nefore you add the sleepers or with the sleepers in place? The latter is hard to imagine because heat would deform the printed parts, wouldn't it? And first way means to work very exactly, maybe with some kind of a gauge?

Hi Volker

Thanks, and good question.

After.

I do have a heat resistant resin, but actually I'm not using it for this (partly as it is light gray).  Using that resin with a colourant would be an option.  I could also use it as is and use a black resin for the frog and guard rails.  Possible, but more complicated, solutions. I'm assuming the rails may take the odd knock in use and thus want to have a dark base colour.

The rivets are light, but the rail is relatively heavy and needs to be heated to make the joint.  It is possible with a soldering iron, but not that reliable.  I'm using a small rivet, but a larger diameter one might be easier and would likely fit, but I don't have them to hand.  You could certainly improve the contact area between rivet and rail by changing the orientation of the rivet and/or refining the closure to the rivet.  It's possible to have the rivet head under the rail (to give a better contact area), but it's very slightly visible and I then need to close the rivets underneath. This is easily done on plain track, but trickier under special chairs.

My method is:

• assemble the turnout
• fit the rivets from beneath
• flux the rivets and quickly fill them with 'solder' (179 degree). Not much heat needed for that.
• flux the rail at each joint. From the underside, with a sharp probe in the resistance solderer press on the rivet and make the joint.  In practice I just do a few at a time. Most of the time the circuit is not made without a bit of pressure.  This way you know you have contact at least - which you don't when just using a heated iron.

The plastic does heat up and even chars slightly on the underside, but once it cools it appears to be fine.  A heavy hand will certainly ruin it, but A few test pieces refined the method. The method is very hot, but very quick.

The turnout is very robust even without soldering.  Test pieces show that some of the solder joints are very good, some less so and some do not flow at all. But sufficient are good to achieve the required result.  It is easier to add a few more rivet locations and live with a few dud connections.

This method is quick and pragmatically seems to work adequately.  If it did not, then there are certainly some things (as above) that could be changed to improve.  I only need 6 turnouts, so I'm inclined to stop at pragmatic success.  If there were a need for many more one would certainly develop and refine methods a bit.

[Lawton Maner]

Can you please post a photo of the rivets you used and the source of them.  One can never have to many places to get supplies.

[Lawrence@NZFinescale]

Can you please post a photo of the rivets you used and the source of them.  One can never have to many places to get supplies.

Dead right Lawton.

A visit to AliExpress.com and search for 'PCB rivets' will turn up images and plenty of sources.  They come in a variety of sizes and there's generally a dimensional drawing too. Note that a quirk of dealing with such sources is that they will often state the material as copper, when in fact it is brass.

[Lawrence@NZFinescale]

I'm into buildings at present...

In the mid C20 many government and civic works were lit by globes with an enamelled 18" diameter shade/reflector with corrugated rim. There were a few at Rewanui (On the Liverpool mine store and the coaling stage (as in the pic) at least).

I've wondered about a method for producing them for a while (and tried etching them 20 years ago).  This time I printed some little press tools and formed some annealed shim brass. They are a tiny detail, but should look good in situ.

It's a good example of indirect uses for the 3D printer.

[finescalerr]

Inspirational. -- Russ