Hi,
I would appreciate some advice from you modelers with some experience with this issue. I would like to incorporate superelevation on the curves on my layout because of the prototypical appearance. My question involves the use of superelevation of curves that are on a grade. Do the prototype railroads utilize superelevation on a curve if it is also on a grade? If so, is it wise for me to combine these two on my model railroad? I model in HO scale and wonder if the combination of superelevation and a grade coupled with a long train (25-30 cars) would tend to pull the train off of the rails, even with cars weighted to NMRA standards. Would I be better off not using superelevation on curves on a grade?
PD
Superelevation is going to lessen your pulling power on the upgrade. In the extreme case, a 90 degree banked curve, the force normal (perpendicular to horizontal) will go to zero. Suince friction is a function of force normal, coefficient of friction, and area of contact, friction will go to zero with force normal. On a slight angle the effect will be significantly less, but if the grade is steep enough that tractive power is an issue, every little bit helps.
At low speeds, superelevation will increase the train's tendency to roll inwards under stringlining forces. At higher speeds, the superelevation will counteract the same tendency. We're talking very high speeds here, IMO. This is true on a grade or on a flat, I don't see the grade making much difference, except when the train's being pulled up a steep grade, but there, the effect is not due to the superelevation, it's due to the grade and covered in my first case.
Can't give you prototype practice, someone else will have to do that. It may vary from line to line, because I know I've seen downgrades superelevated in older steam photos.
Most superelevation on model railroads is strictly for appearance. It should have a minimal effect on the actual operation of the trains. I suspect the real challenge would be the transition from superelevated to normal and just where on the curve that takes place. It could pose a challenge to 6 axle cars and locomotives.
If you think it is going to be a matter of concern, why not build a test-bed? These days you can get used brass track at swap meets for next to nothing. Build a version of the grade you are thinking of and include the superelevation, see if it works. Obviously you'd use N/S rail for the actual layout. The only problem with testing things out with brass track is the reverse Murphy's law effect -- that is bound to be your best track laying job ever.
I use cheap swap meet brass track to try out everything from new methods of ballasting to minimum parallel curve radius for full length passenger cars to testing out S curves in yard throats.
Dave Nelson
tpd0418 wrote: Hi,I would appreciate some advice from you modelers with some experience with this issue. I would like to incorporate superelevation on the curves on my layout because of the prototypical appearance. My question involves the use of superelevation of curves that are on a grade. Do the prototype railroads utilize superelevation on a curve if it is also on a grade? If so, is it wise for me to combine these two on my model railroad? I model in HO scale and wonder if the combination of superelevation and a grade coupled with a long train (25-30 cars) would tend to pull the train off of the rails, even with cars weighted to NMRA standards. Would I be better off not using superelevation on curves on a grade?PD
I tend to agree that it is more for appearances that performance on a model railroad layout. Having said that I did end up superelevating one curve that was on an incline and in a tunnel. I was having problems with 6 axel diesels (only when MUd) going up a curved incline at medium to higher throttle settings. Going down the incline was not an issue. They would cause the front axel to ride over the top of the rail. Superelevating the curve (small pieces of thin styrene under the outside rail) solved this problem. It was quite a proces in trial and error but it worked. If the curve had not also been in a tunnel, I wouldn't have bothered.
Engineer Jeff NS Nut Visit my layout at: http://www.thebinks.com/trains/
Believe it or not - the factors that appear to have the most impact on tendency to "string line" are curve radius and truck rolling friction. Superelevation kept within reason should have little to no impact with a reasonable radius and decent trucks. Heavily weighted cars on steep grades and sharp curves can actually cause more stringlining than the weight prevents. Do not exceed NMRA weight standards, and consistently being somewhat lighter (about 3/4 of standard) and/or less adjustment for length than the NMRA standard can work well with today's free rolling trucks.
As other posters have stated, superelevation on model railroads is pretty much cosmetic. On the prototype, the amount of superelevation required is determined by curve radius and train speed. There is an optimum amount of superelevation for a given curve radius and train speed, but of course the real world would have a wide variety of train speeds, especially if the same track is used for both up and down grade. On even the highest speed prototypes, superelevation is generally less than 6", and often 3"-4". On the model, anything more than the amount needed to give a visual indication of superelevation is wasted. Most model railroaders who use superelevation limit it to about 4 scale inches. A smooth and gradual transition is needed to prevent problems with our rigid, non-equalized, model truck and locomotive frames.
As a side note, long trains on grades on curves of less than 3X longest car length is asking for trouble. Stringlining becomes a real risk for long trains on grades on smaller than 3X radius. That is why I would never recommend a helix with less than 3X radius.
just my thoughts
Fred W
The mistakes that most folks make with superelevation is these: too much, and improper transition.
I have successfully superelevated on grades with curves, and did find that there was some effect on my little 0-6-0. It's traction improved visibly when the curve straightened and it only had to contend with the grade.
There are threads in the archives on superelevation, but a good rule of thumb is to keep the disparity between the two rail heads to something like 2" scaled. About the time you think you like the look, you almost surely have gone too far. Also, use strips of masking tape with staggered overlay to make your transitions. Lay a single strip under the outer tie edges outside the curve that traverses the entire transition zone. Next, add a second layer, but its ends should come no closer than about 1.5" from the first layer's ends. Three to four such layers, with the staggered ends providing the ramp, or transition, should be lots for all the superelevation you want.
I believe that if you encounter any derailments using superelevation, you must not have allowed a sufficient ramp up to the shimmed outer rail. Any more than .020" and not allowing the gradual ramp up at the easment may prove to be trouble. Superelevation on any turn less than 30" radius could also cause problems.
This is no more than .020 throughout a 54" radius
This is the other end of the same turn, note the old dairy siding with the B&M switcher, it was purposely handlaid to have humps and small kinks. The superelevation drops just prior to the turnout as is the easment for the turn itself.
Modeling B&O- Chessie Bob K. www.ssmrc.org
Thanks for the great advice. I know that some superelevation will improve the appearance, but didn't know if it could cause operating problems if used on a grade. I did not want to sacrifice performance for appearance.
From my own experience building a layout with superelevation and grades and 30" minimum radius curves, turnouts and superelevation do not mix. The divergent track heads down when the tangent or outboard track is elevated. Also, due to the fact that most locomotives or cars that have more than three axles coupled together are rigid, the superelevation should be held for a least nine inches on the tangent unless you are also using a gentle transition into the curve. I never had a probelm with string lining on that layout but I also used transitorized power packs that permitted a slow start up. I venture to say that if you have a ten car train that string lines on a moderately superelevated track, flatening the curve will probably have the same effect with only one or two more cars in the train.
I have always used just the amount of superelevation that is perceptible and pleasing to the eye. NO MORE! I always start the superelevation one car length (about 12 inches) in the tangent prior to the beginning of the spiral easement and reach the maximum at the end of it.
I have NEVER had a problem with derailments. Abrupt changes are the cause of derailments. Avoid them like the plague, and that includes vertical easements as well.
I would NEVER use it in a helix. What for? It is only for appearances. I have given thought to a reverse elevation in the helix to offset the stringlining on long trains over a 2 percent grade.
fwright wrote:Believe it or not - the factors that appear to have the most impact on tendency to "string line" are curve radius and truck rolling friction. just my thoughtsFred W
Fred,
I read this entire thread and was surprised to see Brunton not having commented. He'd pointed out in another thread that superelevation was a contributor, ( I remember it as significant or key) to stringlining long trains, especially in the helix. Since he is an ME and has a pretty amazing helix, I've always assumed he was correct. Certainly your two issues of radius and rolling friction would be sure fire causes too.
Interesting discussion. Wonder what the real impact of superelevation on stringlining is? I certainly don't know.
BigRusty wrote:I have always used just the amount of superelevation that is perceptible and pleasing to the eye. NO MORE! I always start the superelevation one car length (about 12 inches) in the tangent prior to the beginning of the spiral easement and reach the maximum at the end of it. I have NEVER had a problem with derailments. Abrupt changes are the cause of derailments. Avoid them like the plague, and that includes vertical easements as well.
Same here. I use it on 30 - 36" curves and 0 - 1% grades with never a problem. I've also tried it on a section of 27" rad at 2.1% with no problems, but that line only gets short trains.
Superelevation may be only cosmetic (except when the kids say "faster faster") but like a touch of makeup on a pretty woman, a little bit of cosmetics goes a long way to create something very nice to look at.
Don't be afraid to give it a try.
Karl
The mind is like a parachute. It works better when it's open. www.stremy.net
That is certainly very sound in theory.
HOWEVER, it just does not happen. I for many years I ran 60 car freights around 30 inch radius curves on a 2 percent grade with 1/8th's inch superelevation and NEVER had a problem.
It just doesn't happen if you follow the advice I gave in my prior post on this subject.
Way back in the dark ages, well, the 1970's, my 4x8 layout had a steep down grade(I never calculated it, but it had to be 5-6% at least. On DC power, the locos even at minimum throttle would race down the sharp curve tossing underweight cars off the rail. BOOM! I raised the outter side of the track(Atlas pier set actually) and was carefull to slow the trains before reaching the grade and no problems.
My new layout, if I ever get to it, will have 30" curves and no grades more than 1.5%. Am I making a scenic marvel? Not really. If I get some good help I may super elevate the mainline as I'd like to see my passenger train run at top speed.
Prototypes...I don't recall any significant grades or very noticable superelevation on the NYC subways(4 years)or MetroNorth's Harlem line(2 years ridership) but it's been a few years.
When we were in the planning stages of our current railroad, we did a lot of research and testing of stuff, old and new. One of the things we looked into was super-elevation. We wanted it, but weren't sure how to do it.
Yeah, we had seen the articles from time to time where someone said they used strip wood or a piece of wire under the outside rail. The problem with that is there is no support under the inside rail, instead the end of the tie is trying to support the rail. If you are spiking on the outside of the rail, either with flex track or hand laid track, it could warp the plastic or snap the wooden ties. In both cases this could lead to a tightening of the gage.
We fooled around with a few approaches and then I stumbled across something interesting in the Walther's catalog. Not under track, but in the details section. Precision Scale makes a line of flex track called "Top Track". In this line is super-elevated track as well as regular track in codes 70, 83 and 100. We ordered a pack (I think 9 pieces.) from our LHS.
It is really cool stuff and shows that someone was thinking when they designed it. The higher side of the super-elevation is acheived by small plastic bumps approximately 1/16"+ high right under the rail. The inside rail is supported by bumps of approximately 1/32", again, right under the rail. Once we saw this we were sold and ordered both code 83 for the main and code 70 for the sidings.
We designed in spiral easements on all our curves. When we cut the roadbed and sub-roadbed we made up templates that had the easements built-in for the radii we were going to use. All our mainline curves (And mainline sidings) are a minimun of 30", larger where possible. From experience with previous railroads we knew we wanted easements, they look better and operate better.
We built a grade scale with a small level attached to do the grades on our risers. After we finished with the risers, roadbed and su-roadbed, we modified the grade scale to use as a jig for making transitions in and out of the super-elevation. We set it so the transition from full elevation to zero would happen in 18". We then laid the trask on the jig ties up, and ran it under a Dremel with a cutter mounted in a Dremel drill press. It worked like a charm. All we had to do was make left and right transitions as needed. We ran both the high and low sides through to get a real transition to zero.
When we laid the track, we started the transition on tangent as it eased into the spiral easement so the super-elevation was completed before we got into the actual curve. Using this method we have no problems with anything tracking off the curve. We run mixed lash-ups of 4 and 6 axle locomotives (Various manufacturers) and have run a Proto 2000 Heritage 2-8-8-2 with no problems. We also have transitions on both up and down grades.
As far as stringlining of causing any additional friction? We run coal trains with live loads, they are loaded with real coal. Yes, they are quite heavy. We run 15-17 cars of triple, quad and bathtub style coal hoppers on grades in the 2% range with a short stretch of 3% in a 30" radius curve. We start and stop them on the grades (With and without a pusher.), again, with no problems.
Operators of various experience, including those with none, have run around the railroad with no trouble.
Now there are some things that go hand-in-hand with this lack of trouble. We were very careful in building the benchwork, everything is level, strong and very stable. The sub-roadbed and roadbed was put down carefully to insure both good cross-level and smooth, hump and dip free roadbed. We even sanded or filled any mismatches in the Homasote to keep things smooth. When we put the track down, we were again very careful, no kinks, smooth transitions into and out of curves. Smooth, even, curves - no tight spots. Everything is soldered and well spiked down.
All this care and work has been well worth it in trouble-free and derailment-free operation of the railroad. We can shove those loaded trains in and out of coal mines and loaders, through #5 turnouts and through yard ladders of #5 turnouts with no problems.
Precision Scale still makes this track (We initially bought it about 18 years ago.) and it is still in the Walther's catalog. There is one strange quirk to this track, they decided to gage-widen it just slightly to help in tight curves. I guess they figured anyone using super-elevated track was doing so with tight curves. It causes no problems and is so slight you barely notice it when joining it to a piece of "normal" track or a turnout. The Top Track has nice tie detail, nice plate and spike detail. There are no holes for spiking, we drilled ours with a pin vise as we were laying it. It's a tough to curve, about the same as Micro Engineering or Shinohara/Walthers track.
The effect of the super-elevation is fantastic, especially when snaking through s-curves. It doesn't show up real well in photographs for some reason, but is almost instantly noticed by everyone in person. This track is a very nice solution to a not too easy to solve problem if you want super-elevation.
I hope this helps you out. Good luck.