S curves are bad


SilverBlade

Member
So as I'm reading and lurking and trying to pick up what I can, I've caught onto the buzz that S curves are a bad thing. What I haven't seen defined is what exactly an S curve IS and when is it no longer an S curve. So in very moron tolerate newbie terms, what actually makes an "S curve"? Is it where ONE car changes turn direction between the front and rear trucks? Or is it when TWO adjacent cars are on opposite turn directions? How far does the turn have to go (in degrees for example)? Isn't any switch to a parallel line sort of a shallow S curve?

And what are the most common consequences? Derailing? Uncoupling? The spoiling of otherwise perfectly drinkable milk?

What clever rules of thumb am I to follow to avoid this frustration. Let's give a nicely succinct synopsis so the next bumblingly twit who types in a search for "S curves" in the layout design thread actually finds something helpful. :)
 
And S curve is a unbroken sinuous wave. Think of a long-framed locomotive coming to a passing siding, taking a sharp #4 turnout, but then as sharply curving back to parallel with the main as soon as the turnout appliance would allow it...at the join, that is. The nose of the loco doesn't do too badly, first diverging sharply left or right, moving through the frog to the join, and then immediately the nose follows the curve as the truck below it pivots sharply to run the tangent making the siding parallel. No biggie there.

How about rearward, though, where the tail of our long locomotive, with its rigid frame, wants to move in concert with what the frame is doing ahead of it under the influence of the re-paralleling front truck? The coupler at the tail end will want to swing in the same direction as the nose of the loco to keep a hold of its mate at the front end of the trailing car. But, the car hasn't entered the S curve yet because it is still encountering the tips of the points rails. At some point, the couplers will lock at their extreme coupler box positions and be able to swing in azimuth no more. The next thing that happens is the trailing car will be levered out of the rails, and probably not the much heavier loco...although a steamer's light tender may be the one to get horsed out of the rails. Its rear end, that is.

Note that there are S curves and there are S curves. The same S curve between two parallel tracks served by a long #8 or higher frog number will be largely innocuous, and much less so the higher the number. It is the lower frog numbers, or S curves that some modelers like to throw into their layout for no real reason other than they think it is cool to have one somewhere along the main, and not because of a turnout. They just like windy track. If they are made with broad curves in the range of 30" or more, it should be no problem. Get down to 18-26" curves, and longer passenger cars, such as the heavyweights, will bind and derail or uncouple.

All it takes to reduce the unwanted effect of an S curve is adding a short length of tangent between the changes in direction. Even on short boxcar length can be a big help.
 
I have seen some nice S curvers on layouts and if done right, there will be no problems. Like the last poster said, as long as it is a broad curve, then you will be alright. It really depends on your personal taste as to if you want it on your layout or not. Me, I do plan on having one on mine.
 
I have a shelf layout with inner and outer loop connected on one side by 2 pairs of turnouts. Much of my rolling stock will make the 1st turn but they derail as they hit the second turn. Short diesel engines make the transition fine but the Big-uns don't like the turns. A smooth sine wave S would probably look nice and work well. Just my experience. Not sure how I will address this, for now I just keep my trains on separate loops and avoid the turnouts.
-Art
 
While I agree with everyone that S curves are bad I would just like to say that I have bachmann's heavyweight coaches (spectrum) and I have four #4 turnouts forming a double crossover with no track in between, just pieced right together and I never have any problems with them de-railing through the turnouts unless I am reversing the ones that still have the horn hooks. Better to be safe than sorry though! Avoid them when you can!
 
While I agree with everyone that S curves are bad I would just like to say that I have bachmann's heavyweight coaches (spectrum) and I have four #4 turnouts forming a double crossover with no track in between, just pieced right together and I never have any problems with them de-railing through the turnouts unless I am reversing the ones that still have the horn hooks. Better to be safe than sorry though! Avoid them when you can!

Those Bachmann Heavyweights have a nifty coupler that twists when the trucks turn. That probably helps them make it through the sharp turns better than my crappy con-cor cars.
-Art
 
S-curves can provide a very nice scenic effect on the layout, as long as the "rules" are followed. I remember reading about S-curves in a Model Railroader many years ago, in one of Lynn Westcott's editorials.

The rule of S-curves he said was this. It didn't matter what the radius's of the curves were, as long as the straight track connection between them were equal to or longer than the length of the longest car or loco that would traverse the curve.

I have many S-curves on my layout, esp through crossovers and such, but I have made sure that the connector track is longer than the longest cars, or locos, that will be using that track. To date any derailments I have had in those curves were tracked down to problems with the car, generally a gauge problem or the truck was mounted too tight to swivel correctly.
 
My current testing track has four Peco turnouts connected to each other without any straight sections inbetween. These are #4 turnouts, 18" radius. I made a quick video of a train going through here at three different speeds - three F45s pulling 21 four-axle cars of varying length-

[YOUTUBE]ov0oipdV-_I[/YOUTUBE]

I can reverse over it ok as well, although I don't dare attempt it at some of those speeds. :D The only place I seem to get derailment problems involving corners is if I put a sharp turn immediately following a crossover piece. Just as Selector said, the car behind the engine gets pulled out of the rails at the frog when the engine enters the curve and the rear swings out.
 
These are two S curves. They are very shallow, but there is no tangent between the three opposing curves. Technically, a tangent between the curves negates its status, so there must only be a single point of tangency. The curve in the middle ground is about 120" radius, and the one over which the H-8 locomotive is moving is about 34" radius.


alleghenynumber2res.png
 
Now thats a S curve. Great job, and I really like the scenery.
 
I couldn’t resist making a S curve on one of my modules today. I only laid the roadbed. The first curve eases from a tangent into a 36” radius then into a 38 1/4 “ radius with a tangent between them of 7.5 inches.
I was inspired of footage I saw from the early fifties of a branch line showing an S curve going tight around a house in a village.
 
Last edited by a moderator:
TrinityJayOne;
That appears to be a universal interlocking (or CP if you prefer). A double x-over. That would not be allowed in the prototype, the signal system would not or should not allow a favorable signal under those circumstances.

Selector;
Which magazine cover did that appear? I'm sure it must of been on more than one since it would be a shame if it wasn't. ;)
 
Thanks for your kind compliment, Bruce. I have no imagery published as of yet, but a hobby magazine has expressed interest in one of mine recently that I believe they'll publish. It will be my first, but not this one.
 
Aw, Selector, yer cheatin'! That hog is one of them ar-tickle-lateds. :D Neat photo!

Seriously, my test for trackability on an S-curve is to run at least two (2) 85' passenger cars over them. I generally keep the diaphragms a bit short, so they don't catch on each other. I'll also test by running six-axle diesels and one of my kitbashed Mantua 2-10-4's or the 4-8-4 through them. Because of space limitation, there sometimes isn't any choice, on passing sidings, for example. Better if you can use at least #6 turnouts. On crossovers between parallel tracks, #6's are minimum, and #8 are better...if you have the room.
 
I have no problems with S curves at all. I have a few on my layout. One, coming onto a passing siding is through #6 turnouts and I have a #6 double crossover that have given me no derailments on code 70 rail. I can run 85 foot passenger cars through them with ease, at a reasonable speed. Although I very rarely run them, a can also run a challenger and my brass Z-5 yellowstone 2-8-8-4 easily.

I have also run long trains (60 cars or more) on both club and home layouts with the Yellowstone and being that the track having been laid with easements the train can easily run throug the S curves at a decent speed. If the track is properly laid, there shoudn't be any issues.

If ther are any joints in the rails in the curves that aren't properly alligned there could be a derailment problem.
 



Back
Top