Ratio: Length of various curves to various straights

[ndbprr]

It doesn't surprise me you can't duplicate a tinplate layout. There are too many variables where the ends of the track join. Your assumption is that every joint on there is a tight connection with the ends of the rail from adjoining rail sections touching each other and alligned properly. I've never seen a tinplate layout where that was the case. In fact the joints are where you fudge things a little for improper allignment.

[Anonymous]

Well, I was considering the center of the rails for my measurement. For my computer simulation, a rail exists as a bezier curve, with each piece of track being determined by a few points.

ndpbrr: Yeah, I tried that early on... no matter how carefully I determine the size of the tracks, I cannot duplicate any of these layouts here:
http://www.thortrains.net/marx/kdlaymth.html

When I try to, I *always* end up with some overlap. I guess what I'm most worried about is that a train layout cannot exist with a splitting diagonal without some bending or stress to the tracks in the real world that's not possible in a computer program unless I make every piece of track act flexible. Not something I wanted to do, initially.

[Anonymous]

The other facor you have is, as mentioned, the difference between the inside and outside rail lengths.

The curve radius is measured to the center of the track, which lacks any physical reference, unless you've got a 3-rail track. If you're making your own curves using flex track, you start out with track as long as the outside rail, and then you cut off part of the inside rail. The amount of track that you've lost in the process is half the length of the inside rail length that you cut off.

As a final note, no matter how precisely you design the layout in the computer, reality when you start to lay track will force you to make a few adjustments. That's how my yard went from 5 to 4 tracks, but it's a better product as well.

When I designed my layout I used AutoCad, and built templates for easements, curves of various radii, etc. It helped for when I translated it to the full-sized drawing. After I built the benchwork, I covered it with drawing paper and drew out the entire layout full-sized by hand. I STRONGLY encourage this final step. It's the final check on your design.

Mark in Utah

[ndbprr]

Surely somewhere in your mathematics in school they taught you that the circumference of a circle is the diameter times pi or two times the radius times pi. You can take that out to 26 decimal places if you wish for each rail since they are a scale 4'8 1/2" apart so they are different lengths.

[dehusman]

You are doing it b_a_s_sackwards.

Design the layout and then bend and fit the track to match design, not bend the track and try to fit the design to the pieces.

Dave H.

[IRONROOSTER]

Not sure exactly what you are doing. If you're showing both rails, neither one is the length of the track piece when straight. The outer rail will be longer and the inner rail shorter. You'll need to calculate each rail separately.
Good Luck
Paul

[Anonymous]

Hi all,

I'm a small time software developer, and I'm trying to write a little program to make modelling my train tracks easy.

My problem is: I'm basically taking straight pieces of track, and bending them, to make curves. I've tried measuring my o-gauge track with a piece of string to get the "real" length of a piece of curved track, and then bend it when I model it in the computer.

When I do this, and make what should be perfectly matched up pieces, they don't fit correctly for building track layouts. (I've also tried the shareware program RTS 7.0, but their tracks seem to have the exact same overlaps and overshootings that mine do.

Does anyone know where I might find the information on how long a curve piece of track is in an "unbent" state? I've been googling for a couple days, haven't turned up anything worthwhile.