Smallest workable S Curve?

DeaconKC Oct 19, 2022

  1. DeaconKC

    DeaconKC TrainBoard Member

    569
    1,320
    21
    Okay, here is a question for youse guys. Watching NPBadlands viseo the other day got me thinking [yeah, I know, scary]What is the smallest reliable distance usable on a S curve? I know this will be dependent on several things, so what are the general ideas and problem/solutions found by you folks?
     
  2. sidney

    sidney TrainBoard Member

    827
    1,001
    21
    nobody likes s curves but there on real rail roads (y)
     
    mtntrainman and DeaconKC like this.
  3. DeaconKC

    DeaconKC TrainBoard Member

    569
    1,320
    21
    Yup, I have tried really hard to avoid them on the layout I'm building right now!
     
  4. pomperaugrr

    pomperaugrr TrainBoard Member

    193
    1,293
    34
    In my experience, the straight track portion between curves must be longer than the longest car or locomotive you plan to have traverse that "S" curve for somewhat reliable operation.
     
  5. mtntrainman

    mtntrainman TrainBoard Supporter

    9,361
    7,703
    137
    Barstow Rick must be offline !! LOL...:LOL::LOL::LOL::ROFLMAO::ROFLMAO::ROFLMAO::ROFLMAO::ROFLMAO:
     
    BNSF FAN and DeaconKC like this.
  6. acptulsa

    acptulsa TrainBoard Member

    3,086
    4,460
    74
    I'm not sure what you mean by smallest reliable distance. pomperaugrr is correct that a straight between the reverse curves as long as your longest car eliminates the problem. At that point, it isn't an ess curve any more. Anything less makes things worse; you're better off running directly from one curve to the other and trying to widen the curvature of them both, as in, use a longer, higher number switches for the crossover.

    I'll bet if you show me the plan I can eliminate some or most of your ess curves. For example, it comes naturally to us to run the main line through the straight leg of the switch, even where running it through the curved leg not only eliminates an ess to the passing track, it makes that siding longer.

    If you mean, what's the tightest radius curves you can jam together back to back, that depends. Are you using a switcher to shunt forty foot cars? Or are you running eighty-five footers through at a scale ninety per?
     
    Last edited: Oct 20, 2022
    DeaconKC likes this.
  7. Inkaneer

    Inkaneer TrainBoard Member

    4,167
    1,007
    73
    The idea of having a straight section of track between the two curves is to prevent a car from having its ends traveling in opposite directions. The straight section allows one end to be on a curve while the opposite end is on a straight section. The effect is lessened by using easements between the two curves but space in our model world may not allow such easements, so the straight track section is used.
     
    x600, pomperaugrr and DeaconKC like this.
  8. acptulsa

    acptulsa TrainBoard Member

    3,086
    4,460
    74
    Well, that is part of it. But the biggest problem happens when a car on a right hand curve is coupled to a car on a left hand curve. At that point the centers of the two cars are some distance apart.
     
    DeaconKC likes this.
  9. DeaconKC

    DeaconKC TrainBoard Member

    569
    1,320
    21
    I actually have already gone through my planned layout and got rid of the 2 offending curves after seeing the Badlands video. It just got my 4 brain cells wondering [wandering?] about how someone would avoid that problem. My layout is going to be limited to the 65' and 70' passenger cars, 2-8-2s and a set of Alco PA/PB units as longest stuff on there.
     
  10. Inkaneer

    Inkaneer TrainBoard Member

    4,167
    1,007
    73
    The problem is exacerbated with tight curves and long cars that are traditionally found in our model worlds. Common areas where an S curve is found is in a crossover from one track to another or a siding where the siding is parallel to the main. Long cars and #4 switches don't always work and play well together.
     
    DeaconKC likes this.
  11. DCESharkman

    DCESharkman TrainBoard Member

    4,012
    1,430
    70
    It all depends on the radius of curvature on both branches on the S curve as well as an easement track before and after the the curve. If the radius is large enough there may not need an easement track either. When you get to 36-48 inch radius of curvature, there are no issues with S curves.
    Before you jump all over me, I know very few people can have curves that large, so that is where the easement curve on both ends can make a difference. And there is nothing to say you can't separate the curves in between with an easement section of track as well. Just tinker and experiment and you get them to be reliable.
     
    mtntrainman, DeaconKC and acptulsa like this.
  12. BigJake

    BigJake TrainBoard Member

    2,149
    2,803
    57
    The N scale Unitrack single crossover has an S curve between the two adjacent #4 switches' diverging routes, which are 19"R curves. Pretty much anything can go through that. Whether it looks good, well, that depends a lot on the viewer looking at it, and from what angle.

    I have often posted that I use a formula, borrowed from calculating parallel resistance, to calculate an equivalent severity transition of an abrupt change from straight (tangent) track to a curve of given radius Req:

    1/Req = 1/R1 + 1/R2

    Solving for Req:

    Req = 1/(1/R1 + 1/R2)
    Per that formula, the S-curve combinations with >= 9.5" Req in N scale Unitrack are:

    28R/28R (14 Req) same as Unitrack double crossover
    28R/19R (11.32 Req)
    28R/15R (9.77 Req)
    19R/19R (9.5 Req) same as Unitrack single crossover
    28/13.75R (9.22 Req)
    19R/15R (8.38 Req)


    Note that if one curve is actually straight, then it's radius is infinity, and 1/infinity=0, so the formula reverts to Req = R.

    Finally, as noted previously, "straight" track is not effectively straight until it is long enough to hold the car traversing it. However, even a ~2.5" length of straight in between opposite curves is enough to allow most cars to easily traverse,
     
    DeaconKC likes this.
  13. BarstowRick

    BarstowRick TrainBoard Supporter

    9,119
    4,145
    140
    I saw this, George!

    Here we go again. You just had to ask. Didn't you Deacon?

    A 24" radius curve is about as tight as you want to get with S Scale.. I hope that helps.

    The sectional track I got with my S Scale Heavy Weight Passenger trains exceeded the 24" radius. I could not build my layout on a 4X8 sheet of plywood. The curves hung off the layout by about 4"s.

    For a boss of mine I was able to use a type of Flex Track and used a minimum of 24" radius curves. When you look at the older S Scale American Flyer switches, those curves in the switch are tight. You might get tighter curves but then (#$^!@#%&&*&^%$&*^%) you might not like the results. I went looking for S Scale pre=fab switches in the #6 arena. Didn't find any. Era 2000 to 2010. Some stuff available but not much. It's better today.


    Edited add on: Oh boy, did I blow it. I thought you were asking about S Scale. Don't read so fast..

    N Scale S-curve. My advice is still good and the dimensions work. It will make an awesome S=curve.

    Still...Things to do.
     
    Last edited: Oct 20, 2022
    BNSF FAN and DeaconKC like this.
  14. NtheBasement

    NtheBasement TrainBoard Member

    376
    446
    19
    Regarding the OP question, I think it depends on whether you have truck-mounted or body-mounted couplers. AFAIK truck mounts don't have any problem on ess curves, but worst case is you need a straight transition that is one truck-length long for the car behind the body-mounted loco.
     
    DeaconKC and BarstowRick like this.
  15. BigJake

    BigJake TrainBoard Member

    2,149
    2,803
    57
    I would put that minimum straight transition length at the longest coupled length between the far axles of any two coupled trucks. Keep in mind that the length between coupled trucks on many railcars with long overhangs can get pretty long.

    Curves create lateral torque about the coupler-equipped truck pivots, dependent upon coupler shank lengths, and the resulting distance between coupled truck pivots.

    Especially in pushing operations uphill, torque on the truck pivots is exacerbated by abrupt changes in track radius (tight S-curves being the worst case.) The direct torque applied by the truck mounted coupler on the truck pivot causes a comparatively short wheelbase truck's wheels to jump the rails, especially on railcars with long overhangs and coupler shank lengths to match. Body mounted couplers apply zero torque on the truck pivots.

    Also, most N scale locomotives offered today use body mounted couplers. When pushing on a truck-mounted-coupler-equipped railcar, the most torque is applied to the truck on the railcar coupled to the engine. Some users use "transition railcars," with body and truck mounted couplers on opposite ends, to transition between the locomotive's body-mounted couplers and the truck-mounted couplers on rest of the railcars.

    Generally, pulling favors truck-mounted couplers, especially in short radius curves and short transitions, while pushing favors body-mounted couplers, especially up grades. Finally, abrupt changes in curve radii generally cause problems before absolute radii do.
     
    DeaconKC likes this.

Share This Page