Le Corbusier wrote:Really not sure what I'm doing ... but following your advice my set bends and knuckle bends are definitely working and I am definitely using Exactoscale chairs and definitely not using 'the proper conical press tools' ....
If there is a knuckle bend in inclined rail, the head of the rail will be going round a larger radius than the foot. That means to turn through the same angle the head needs to be longer than the foot. That means the head needs to be stretched relative to the foot, or the foot compressed in length relative to the head. There are two ways to do that. 1. make the bend using a conical mandrel while firmly holding the rail inclined and in line. 2. make an ordinary bend and then bolt each side of it down flat with the rail inclined. The prototype uses both methods, using big lumps of cast iron, shaped spacer blocks and bolts.
On the model, the chairs are made of tiny bits of soft plastic. The rails are hard metal. Guess which one wins in a fight? I'm willing to bet that where you have made knuckle bends your rail is actually vertical.
The same applies to curved plain track. If the rail is inclined, the head must follow a larger or smaller radius than the foot. It must therefore be stretched or compressed in length relative to the foot. That requires significant force, which can only be achieved using heavy fixings. A flimsy plastic chair can't do it, except at the gentlest of radii. When curved the rail will simply force itself vertical in the chairs. One of the first makes of inclined flexi-track claimed that the track was self gauge-widening on curves for this reason. In practice the vertical rail is predictable, but the effect on the gauge between the rails is not. It depends on the exact fit of the rail section in the chairs.
re. self-guiding of conical wheels. Even on the prototype this works only at gentle radii. At sharper curves the wheel is guided by the flange, which runs hard against the rail head. Hence the wear and side-cutting of curved rail, and the need for flange greasers. Were it not so, there would be no need for the wheel flanges. On a model with sharper relative curves than the prototype it has been shown that the coning has minimal effect. Try running your layout with flangeless wheels and see what happens.