The Dreaded "S" Curve

[Don Gibson]

PROBLEM is one car end going one way while other end goes another. Trucks are affected - as are couplers and car diaphrams.

IN THEORY a piece of straight as long as your 'longest' car is ideal. Articulated Cars with common  (shared) trucks, centered (sometimes called a 'Talgo' ), don't have this problem .

IN PRACTICE:  Trucks are mounted in from the ends, so the distance between pivots is more accurate.

PREFAB Crossovers using NMRA specs generally have about 4"-5 " of straight frog-to-frog, so theoretically are fine for 40' or 4" cars, but work with longer, depending on trucks, pivot freedom, and speed taken.

ALL NMRA (numbered) turnouts are supposed to have straight beyond the frog, wheras European NEM  (and Atlas 'Snap's witches are designed to curve. WORST case is two reverse 'curved' sections mating directly into each other.

OTHER: Cars follow a curved track by flanges pressing against one rail.  A sudden shift to the other rail is disturbs the 'flow.' I had a 26" LH curve feeding into a switch with a RH #6 crossover, which caused derailments. Changing  the crossover to a LH cured it. (Dynamics of an 'S' curve).

I agree with advice given to  add some straight transition between curves - even at the expense of tightening the curves. WHETHER it's 4", 6", or 9", Most of today's equipment will run on 24"r. so it's up to you.

[twhite]

I have an "S" curve on the return loop on my MR.  The radius is 34", and I installed a 4" straight section in between (all I had room for because of the way the curve comes back into the main), and no problems at all, even with 85' passenger cars.  I think it really depends on the tightness of the radius as to how much trouble you're going to have with cars on the curve.  I know I certainly wouldn't recommend an "S" curve without a car-length straight track in between if you were running anywhere from 18 to 24" radii. 

Tom

[gkhazzard]

I think the best advice you're gong to get is some that a couple have laready given and that I will reiterate... test it.  If you have a definative era set, then it's pretty easy to determine the largest available cars and locos for the time period.  Grab some, and run them through a test set up and see what happens.

[mike_ruby]

I have two spots on my layout where I am faced with "S" curves. The radius of two curves is 31" and 34". My question is at what radius in HO do "S" curves no longer present a problem? Or is there such a magic number? Thanks

Brent

I have had the same problem on my layout, almost the same radii, 32" and 34". They were slightly larger, but I found that my 69' flat cars (with fixed coupler pockets) would not go around the S-curve. I decreased the radii to 32" and 34" which gave a 6" tangent between the curves. No problems with the stock now.
If using shorter cars you may be Ok.

Mike Ruby

[canazar]

31 and 34 radius are pretty big so it sounds like you have some room to work with.  if you all ready have it laid, test everything you got and see if it clears.  I have see some S crurves that work.

If track hasnt been laid yet and you are workign off of mock up or drawings, I would try and take out some of the middle of the S.  Even if you have to lose a little in the 31 and 34, it could help equipment run better.  

If yoru running shorter rollign stock, such as 40-50 footers and short steamers and G's and F units, Id say you would be ok.  (I have seen worse work)  but if you want to run bigger power and the killer, passeneger cars, I would definatly look into easing out as much of the "joint" of the S curve.

If you are in a bad spot and have to work with-in limitations and are trying to get the most of out of your space, I would at least lay the track and test it with the biggest stuff you got.  Could be worth the peice of mind.

[rerailer]

Remember to have at least one full car length of straight track between the 2 curves. 

[tomikawaTT]

The problem with S curves is that the end sills of coupled cars are pushed in opposite directions.  If the coupler side swing can't accept the offset, one of the cars will be forced sideways up and over the rail.  So the radius depends on the rolling stock - short for ore jimmies, longer for 40' box cars, much longer for auto racks and full-length Pullmans, humongous for two NYC Niagaras coupled tender-to-tender.

If you absolutely, positively HAVE to lay an S-curve, the safest way to prove the concept is to do a test-lay with a couple of pieces of flex track on a nailable tabletop and push your rolling stock through it with an 0-5-0.  Then remember that you will have set a length-of-car limit.  You won't be able to add longer rolling stock in the future.

I have encontered coupler side-thrust problems between long and short cars on short-radius curves that weren't S-curved, so I tend to take the conservative approach.

Chuck (modeling Central Japan in September, 1964 - avoiding S-curves)

[Mark R.]

The secret to "S" curves is the tangent between them. You don't want the train leaving one radius and immediately slamming into an opposing radius.

If you were to decrease both of your radii by one inch and create an easement between the two opposing curves, it will perform considerably better than your current radii running straight into each other.

Mark.

[BATMAN]

I have two spots on my layout where I am faced with "S" curves. The radius of two curves is 31" and 34". My question is at what radius in HO do "S" curves no longer present a problem? Or is there such a magic number? Thanks

Brent