Going Around The Bend!

One of the things I enjoy about this hobby is not just the building and running of trains, but the little engineering challenges that go along with it.

I don't profess to be an expert when it comes to hand laying track, far from it. In fact I would consider myself a beginner in every aspect of it, and one of the things I have been struggling with is the area of laying a consistent radius curve.

I know of various tools one can use including the Tracksetta gauges, but it would seem to me that you are still trying to pull a basically straight length of rail around a curve that will be in tension and try to pull straight if it can.

Metal in general is a funny material and reacts differently when subject to heat and cold and other forces. A piece of code 55 rail is quite delicate, and yet has quite a resistance to being curved and maintaining that shape, and will pretty much spring back straight if given the chance.

Yet with some very subtle pressure, if you run it through a set of rollers it will form a curve and hold it without any trouble at all. I'm not sure of the technical reasons why this happens, but I know from using similar methods for bending much larger pieces of steel that it just does.

Armed with this, I figured it wouldn't be too hard to build a small scale rail bending apparatus. I had a look around for some small roller bearings, and came across some that I think are used in remote controlled helicopters, described as follows:

Esky Belt CP 000372 (EK1-0551) - Bearings 4x7x2.5

I believe EK1-0551 is the actual part number, and the the 4x7x2.5 relate to the size of the actual bearing, 4mm inside diameter, 7mm outside diameter and 2.5mm wide. I bought a set of four off ebay and paid $5.05 delivered.

I used a piece of 32x20x1.6mm aluminium angle cut in two 60mm lengths and clamped them together so they formed a "U" shape. I drilled two holes and slotted another hole to give adjustment to the position of the third bearing.

I found that a set of 8BA bolts (pretty sure they were but I bought them about 25 years ago!) had a head diameter that gave a neat tight fit inside the bearings which was handy. I then drilled the holes in the top piece of aluminium larger than the bottom piece so that I could use an 8BA nut that would allow me to adjust the bearing height, and then another 8BA nut under the bottom piece of aluminium to secure the bearings in place.

The slot for the bottom bearing allows a different amount of tension to be put on the rail to give a very subtle or quite dramatic bend in the rail. With a bit of pressure the rail will bend into a 10cm radius and hold it perfectly!

You must push the rail through the bearings to have the best effect as well. Drawing it through in a pulling motion tends to give a less consistent result as pulling the rail past the last bearing at any angle bar straight changes the tension on the rail and the curve.

The prototype is rough and ready, and is not easily adjusted for tension, however it proves the general theory and the suitability of the bearings I chose, and I can now design and build something a bit more flash.

Ideally the two static bearings could be a little further apart, but the big gain will come in the easy adjustability of the third bearing that determines the degree of curvature.

I have another use for these bearings as well but that will be for another time once I've had time to experiment.