The video makes it!

Wow.

Using video to bring the car/curve performance to life is just amazing.

Awesome

I agree the video really makes it easy to understand where the problem occurs and what the end result is.

I really enjoyed this article as I am battling some S-curves through turnouts in my track planning.  After seeing this I think I can accept the odd S-curve as long as I keep in mind how car length relates to the radius of the S-curve.

Thanks for a great read Tim!

Great article, very timely..

Tim, thanks for the perfect article for my needs right now. I am working on the design of my next model railroad and one of the things I have been thinking about is the role S curves play on my layout. I think your article did a fantastic job sorting fact from fiction. I am a lot less worried now about putting a turnout diverging in a direction opposite the curve preceding it when the preceding curve has a broad radii. Most of these cases on my layout are either very broad cosmetic cuves or terminal/spur situations.

I thought the videos were exactly perfect as a companion to the article.

As I read the article I started to think that there must be an underlying ratio of difference in Radii to car length  that could be used as a tool in evaluating whether an S curve is going to be an issue or not... and probably a base line radii that can be considered tangent for practical purposes.

Chris

S Curves

Hi

I was intrigued with your article on S bends.  I first studied this problem while designing a layout as a kid.  Later I studied this further while developing a road design programme (COGO) back in the sixties and while designing various roads since.

Curvature is the parameter used in preference to radius.  Curvature is the inverse of radius - zero curvature is a straight line which is a curvature of infinite radius.  Transition curves are built into real track for a number of reasons including, avoidance of shock loading at the beginning of curves which cause the track to drift increasing the curvature at that point, a length over which cant can be built up, reduction of the misalignment of the ends of stock during changes of curvature.  The effects of changes of curvature can also be reduced by inserting intermediate curves. 

An S bend is a sudden reversal of curvature.  The effect depends on the curvature involved so an S bend from 3ft to -3ft is equivalent to the start of a 1'6" curve.  The S bend can be mitigated by inserting a straight between the two curves.  The effect of the straight increases with length until the length is the length of a coach less the overhang from bogey centre to buffer face - i.e. the point at which one coach is effectively on the straight while the adjacent is effectively on the curve.

Crossovers consist of two turnouts facing each other.  In fact the turnouts are rarely curved throughout.  Frequently the curve finishes just before the frog and the straight may even include the entire length of the check rails.  Furthermore in the interests of compatability PECO turnouts are built with standard turnout angles and varying radii and overall length.  This means that the length of the straight in a crossover is considerable and is greater for the tighter curvature product.

Thus the worst S bends occur on plain track if the designer omits a transition or straight section of set track between the reverse curves.  This can be mitigated considerably by inserting said section and this can generally be achieved with minimal changes of the overall geometry of the layout.  When transition curves were first introduced on railways back in the 1880s-1920s a number of lines were realligned in order to allow them.  This had to be done carefully within the limitations of existing structures.  A transition curve can be allowed for during early design by offsetting the curves, this means that the distance between the centres of curvature of adjacent curves is greater than the sum of the radii, but not by much.

British stock have buffers this makes alignment of the ends of stock more critical to avoid buffer lock which inevitably results in derailment. 

Archie

My experience differs slightly from Tim's

I have found that engines pushing strings of cars (more than two), have more problems with s-curves than Tim showed. Now Tim's probably forgotten more about handlaying and trackwork than I'll ever know, to be sure.

But I honestly have seen a difference in the build-up of truck skew, coupler thrust, and off-center forces in actual use over multiple cars (especially of mixed lengths and coupler mounting) than what is demonstrated by gripping one car in the center and pulling and shoving a second car with it.

Others mileage may certainly vary. But I hope that MRH readers won't consider every S-curve a non-issue because of this article, setting themselves up for potential reliability headaches later, especially with trackwork and rolling stock that's not up to snuff.

Bottom line, I liked the concept of the article, but my own experience with actual trains of mixed-length cars being shoved by model locomotives suggests s-curves are more of an issue than the videos suggest.

Byron

S curves and turnouts

Boy, this format with videos sure does make for easy comprehension!

What I don't remember is what is the radius curve if you put - say,  in Ho, a #5 left aligned into a #5 left as in a passing track?  

every picture tells a story

Thanks for using so much video in the article, Tim - very enlightening to see it in action. There's no question that it pays big dividends to test things like curve radii and track alignment prior to constructing your dream layout.

Re: Byron's caution against taking this article as a panacea; using an engine & throttle instead of Tim's hand would probably make some measurable difference and really wouldn't be that hard to manage in a test case. Surely, if anyone were contemplating a complex area, some testing (and practise laying track) would be in order before the "main event".

Made me think a little....

I am wondering if the Bronx Terminal plan has even a little bit of easement in the curves?  I would think even a really small amount might make a huge difference in how "bad" the S-curve is.  I also expect (without any basis in fact) that the slow speeds in such a place help as well, as there would be less jerking of the couplers.  If I remember right Armstrong used the term "lurch".

Sorry to bring this up at this late date, Bryon

Sorry to bring this up at this late date, Bryon, but as I was watching Tim's latest video in HD of the Bronx, I noticed his KaDees all had their pins removed.

Have you noticed any significant lateral thrust from the pins? I was wondering if that may be adding to the S-curve issue. More likely than not they probably don't, but I was wondering about that combined with overtightened coupler screws, as I do see that problem when pushing with the Sergents (being too tight in the coupler box that is).

Kadee pins

Normally (if installed properly) the pins of Kadee couplers shouldn't be touching. That is the shouldn't be installed facing exactly fore and aft but at an angle off to the side a bit.

The biggest source of derailments on the Bear Creek (well one of the biggest) is Kadee couple pins that are hanging too low (usually because there's something wrong with the coupler pocket allow the couplers themselves to be drooping. If a low hanging pin encounters a closure rail in a turnout or a grade crossing it can easily derail a car.

But I've never seen any problems with the pins pushing on each other. Not to say that it *never* happens...

Cheers,

Charlie

bending pins

Charlie, maybe a good article idea would be how to set your Kadees in the pockets? Not that I'm using them anymore, but I used to hate bending pins, even with the Kadee plier, as I always felt I was going to damage the coupler or pull it out of the box.

Seeing it in action (video) done by somebody who knows what their doing, might be useful!

Kadee pin bending

That's a good idea. I've not had any problems adjusting coupler pins but perhaps an article on coupler selection and height adjustment using the Kadee gauge would be good.

Cheers,

Charlie

In fact, I'd like to see a definitive coupler article

I'd love to see a definitive couplers article that covers Z-G, with a couple pages on each scale. To do that we'll need each of you who model in those scales to take some photos and send us a couple page piece on what you're doing.

Do we have any takers? Pass the word - send us your coupler experiences - especially if you model in something besides HO or N.

The Coupler Issue

Joe, you might consider a comprehensive Coupler issue - THE issue that covers all couplers, all scales, all time...well, maybe not THAT expansive, but you know what I mean!!

If you do, make sure you get an editorial in there that mentions the future of the coupler, in your opinion, especially in face of the MTH Remote Operating Coupler that will be on their HO SD-70ACEs and the Remote Operating Coupler that will be on a future run of the Kato N 2-8-2 Mikado!

Oh yeah - and maybe an article on tender couplings and drawbars!