Transition curves...

[Peaky 556]

I know there is a lot of experience out there, and I have laid gentle flowing curves with seamless transitions of radius in my limited track laying so far (basically all done by eye), but I am presented with quite a tight space for a curve of around 80 degrees.  The limits of the land dictate that I can either fit in a constant radius of 4.8m, or try and be a little cleverer and have a short stretch at beginning and end with say 6m radius, and then a tighter bit of 4.5m in the middle.  Obviously there are an infinite set of combinations I could try, but I didn't really want any part of my circuits to be less than 4.5m radius.
I think I know the answer, and am prepared to try and install it (more challenging!), but would the consensus be to incorporate the transitions?
Thank you,
Tim

[IanT]

They are a good idea in theory Tim but whether you'd notice any improvement in practice I'm not sure. However, if you are willing to go to the extra effort, then why not - the appearance of the track would be better I think, regardless of its practical impacts.

I did some work in these areas some time ago when trying to write G3 'standards' for the Society. I looked at both European & US railway modelling standards for reference at the time. I've just quickly checked and can't find anything about transitions - but I did find a draft 'Clearance' document for G3. With regards your tunnel enquiry, I've also got NEM 105 - 'Tunnel Profiles for Standard Gauge' (1987) as a PDF. So a bit long in the tooth now but I doubt the basics have changed too much - although you should check.

PM me if you'd like the '87 version and I'll email it to you.   

Regards,

IanT

[Andy B]

Tim,
Strictly speaking, what you're describing isn't a transition (a spiral), but replacing a single radius with 3 compound radii, each of which will need a transition between them and further transitions at their ends.
Even at 80 degrees it isn't that long a curve, so any 6m sections would be quite short.
Why not just keep the main curve at, say 4.7m with a generous transition at each end?

Andy

[dajo]

I have been very, very, pleased that I used transitions on my railway.  Transitions look good, and the smoothness of the trains entering and exiting a curve is a very pleasurable sight.  I would not do it any other way in the future.

I put a lot of effort into writing computer code to generate Euler spiral transitions, with the ultimate practical output of full-size paper templates.  Design experimentation over the parameters (infinite set of combinations): length of transition, base radius, rail length, missing the tree, etc., is easy with this code.

If you have any interest, I shall be very happy to plug in your parameters and send you computer files (.dxf, .pdf, .jpg, your choice) with the results.  It is hard for me to believe that they would not be useful to you.

dajo

[Andy B]

I just do the design in Templot - all the hard work has been done thanks to its developer, Martin Wynne.

Andy

[Peaky 556]

David,
I really love smoothly transitioning curves too, but unfortunately in this case I am constrained by walls, hedges, trees and a prized shrub!  My options really are very few.  Thank you very much for your offer to design it properly for me and achieve the technically best optimum curve within the constraints.  I can see this would be beneficial were I having trackbed sections numerically cut, so that all I had to do was to lay the tracks to follow the predefined curve.  What I had failed to say though is that the track will be laid on a concrete trackbed at the bottom of a trench, and my civils skills (mainly a lack of) will not do your calculations any favours!  I shall therefore decline your kind offer with thanks!

Andy,
I do like your pragmatic approach of blending the start and end into a transition of some kind, feeding to a tighter curve in the middle.  I plan to get some track curvature templates cut for the mid section and dig the trench using them as guidance.  I think at the end of the day it will have a gentle start and finish (because it's difficult to do anything else unless using "set-track"), it will look ok by eye and not have any obvious nodes or kinks, and generous side clearances will keep the stock apart.

Ian,
Appearance of the running tracks, although hopefully elegant, will not be of any subsequent importance because as you may have guessed by now, all of this curve is to be contained in a tunnel!  Hopefully the cats/rabbits/foxes will be impressed 😉.  Regarding the NEM 105 for tunnel profiles, again my space constraints will limit me to something like the Berne loading gauge, which ought to be perfectly adequate for most Gauge 3 situations.  Thank you anyway.

I appreciate this forum as a sounding board to develop my own ideas as well as learning from others, so thank you all!

Regards,
Tim

[MikeWilliams]

Tim,

I've not chipped in because my track isn't running yet and I have little personal experience, but you mentioned "I plan to get some track curvature templates cut for the mid section".  What I did was have some steel plates cut for the main radius (15ft for me) and the short length acting as a transition (30ft in my case).  They are 63mm wide, almost 6ft long each and with holes spaced down the centre so that, with pegs, they can be joined at any point required to give quite a long section.  I'll send some pictures if that is not clear.

Mike

[Peaky 556]

That sounds brilliant Mike. Pics much appreciated, and whether you collected from a local laser cutter because of the difficulties in posting such long items.
Thanks v much,
Tim

[MikeWilliams]

I always collect from my lazer cutter because he lives down the road!  Not the best pictures taken some time ago.

[Peaky 556]

I like those very much Mike; can I pinch your idea please?  Can I further ask for dimensions please, such as thickness, hole size and spacing, and whether if you made them again after your first use that you would change anything?  I may as well make mine compatible with yours; you never know if you'd need to borrow in the future!  My radii will probably be 4500, 4700, 4900.  At risk of admonishment and with no offence intended, I'd say in white-man's terms that's 14' 9", 15' 5", and 16' 1".  I'd say a couple of each size would give good flexibility, if not too costly.
Thanks, Tim

[MikeWilliams]

No problem Tim.  I'll email the CAD file to you.  Your radii all sound a bit similar.  Do you think you'd be able to see the difference between them?  You are welcome to borrow them (if Boris allows) and lay them out on site.  I'll not be using them for a couple of months or more.

One mistake I made was economising by using plain mild steel and then leaving them on the verandah of the summerhouse and they have a light coating of rust which doesn't really matter but hey are now dirty to handle.  Should have used stainless.

Mike