Not sure if this has been discussed previously, but in contemplating building a rack (cog) based project how we get the correct behaviour of any traction rolling stock, particularly if improvising as might have to be the case? Some rack railways have gradients as steep as 25% (1 in 4) so releasing the brakes is simply going to result in an uncontrolled "whoosh" down the hill, whereas anything with a semblance of real physics is not going to climb anything greater than around a 5% (1 in 20) gradient which is close to the 1 in 17 or thereabouts that is the maximum for adhesion worked lines in Switzerland.
The same principle goes for lines which included inclined planes to link the top and bottom sections of conventional railway, usually rope or chain operated (of which there were several in the UK).
The obvious solution would be that any trains on a purported rack section would need to be driven in DCC mode or even have the AI operate them, assuming even then they won't protest at the extreme demands on the traction and braking algorithms.
The same principle goes for lines which included inclined planes to link the top and bottom sections of conventional railway, usually rope or chain operated (of which there were several in the UK).
The obvious solution would be that any trains on a purported rack section would need to be driven in DCC mode or even have the AI operate them, assuming even then they won't protest at the extreme demands on the traction and braking algorithms.