Superelevation question

[Puddington]

This Canadian Thanksgiving weekend Sam and I are working on his layout and we're laying track (yaaaa!) We want to slightly super elevate the curves, @ 28" radius. I haven't done this in ages and seem to forget the height of shims used... 0.010 ? 0.005... ? Any wisdom available... don't want massive modern mainline curve super elevation but just a slight elevation as found on a secondary line of the 70's....

Thanks in advance for any help.

[davefoxx]

Puddy,

I think that you'll want to go taller than 0.010", or else the superelevation won't hardly be perceivable.  Besides, 0.010" only scales out to 1.6".  If you were looking for 4"-5" of superelevation, which, to me, is where it starts to become perceivable, then you may want to raise the outer rail by 0.025" to 0.03125".  I use five to six layers of masking tape for my superelevation, and it's gives a subtle effect without causing operations problems (my minimum mainline radius is 12-3/8").

Hope this helps,
DFF

[C855B]

What Dave left out is the masking tape method also makes smooth transitions easier. The bottom layer is the longest, then shorten each subsequent layer by the length of your horizontal transition ÷ (the number of layers – 1).

[randgust]

I've put regular stripwood (running with the rail) under the outer tie edge between the track and roadbed to test superelevation appearance and operation.  If it doesn't work, it's really easy to change.   If it does, then you ballast right over top of it.    But by all means put it in and test it - extensively.   It can do some really interesting things to certain locomotives with long wheelbases and while it looks great, your chances of stringlining a heavy train on a sharp curve go up drastically.

I superelevated my curves many years ago with very few issues, but that's also when flanges on locomotives were generally deeper.  Its only been since finer-scale flanges came out that I found I had a couple places that needed adjustments.   The other real 'glitch' has been on 85' intermodal flats with body-mounted couplers, those have proven to be highly sensitive to stringlining on a superelevated curve.

[conrail98]

The best kind I've seen done didn't use shims, instead it used masking tape. Take a strip and lay it out under what will be the outside rail from the start of the curve (I believe they did it after the easement, but I could be wrong). Then go your longest car length into the curve and add another layer until you are at the other end at the longest car length, i.e., the strip is 2 car lengths less then the first strip. The few people I saw do it did it up to 5 layers, each layer starting one car length later then the last (and ending one car length sooner) but you could go less depending on how much you wanted. Doing it this way instead of shims is you have a transition into the super elevation which is a little harder then the shim method,

Phil

[randgust]

The only reason I liked shims is that I could tweak it without actually removing the track, particularly reducing it.

Yeah, the real key is the 'runout' of superelevation back to the normal track surface, and how that works out to the spiral.   On a real railroad design, this is far more precise and mathematical than you'd imagine.   FRA track standards are now pretty brutal on how this has to be accomplished.   On the model, flex track is stiff enough that it tends to come to its own runout if you let it, but the length of those runouts is still longer than you'd think.

The entire design of those spirals and superelevation is incredibly significant.   We did one feasibility study of a high-speed rail line in a state not to be named, on a congressionally-designated route not to be named, and current FRA track standards for curves and spirals absolutely prohibited running above about 50mph due to the abundance of reverse curves running through a river valley.   While the 'historic' class 1 in steam days scheduled track speeds through there of 65 with wicked superelevation, the maximum overbalance limits today prohibited it.  You couldn't get out of the curve, flatten out, and reverse direction in the horizontal distances that were there.   So much for a 'high speed' route.

I've seen lots of incredible superelevation in the late steam days, particularly on the Erie in my area.  There was one wicked curve out where I lived that had a 60mph RESTRICTION on it, and to do that, it had at least 6" on the outside rail, as the rail base was visibly above the head of the inner rail.  During Conrail years it was still there, only the track was being run at 40mph (FRAIII) now and the low rail was just getting beat to pieces running that slow with that much superelevation.   Conrail had to take out a lot of superelevation at various places due to that.   The higher center of gravity on steam and the higher speeds of that era left some memorable designs, let me tell you!

[Kisatchie]

You mean THIS isn't super elevation?

https://www.google.com/search?q=chicago+el&client=firefox-a&hs=avI&rls=org.mozilla:en-US:official&source=lnms&tbm=isch&sa=X&ei=u_U3VJmHBOTisASin4GwDQ&ved=0CAkQ_AUoAg&biw=1074&bih=812#facrc=_&imgdii=_&imgrc=JPMspqrKn4I8HM%253A%3BJ4Ooba8xraOMfM%3Bhttp%253A%252F%252Fcollegeofcomplexes.homestead.com%252FCTA_el_train_new.jpg%3Bhttp%253A%252F%252Fwww.getawayhostel.com%252Fmoving-to-chicago-2741.html%3B500%3B375


Hmm... no, Kiz. That's
ultra elevation...

[GaryHinshaw]

I used .020 x .080" styrene strips cut into ~2" lengths, placed under the outer rail.  I transitioned into the curves with a few lengths of .010 and .015" stock for good measure.  It has worked very well, but it is very subtle and I think going as high as .030 would have looked more dramatic, and would probably have performed fine.

Happy Thanksgiving.

[randgust]

I can't remember what I put under it for depth, but that's a 13" curve where 5616 is sitting and it's the heaviest superelevation on the entire layout.  I just love the looks - creates a lot of drama.   It's dead flat - that's another thing -  putting superelevation on grades certainly makes it work differently.

[JSL]

There is still a post on CSX Dixie Line's blog, that shows how he did his superelevation. I just checked and all pictures are still present. I sure do miss Jamie's blogging. RIP.

[randgust]

One rule of thumb that I think is practical - I've seen a lot of stiff frames out there that just won't 'twist' much going into superelevation.   If you have a .022-.024 flange (which is pretty much standard today) you darn well better not twist the track more than .010 in the length of your longest locomotive outside wheelbase; say maybe 65', or risk lifting a truck off the rails.   It tends to be the trailing truck on a unit that will lift and climb over going into the curve.   So you'll need a runout of about 120 scale feet to go into a .020 outside rail raise.   That's the kind of field testing you'll need to do yourself.  I know I was REALLY surprised to see how stiff my six-axle Katos were, but the flanges were sufficiently deep that they don't derail but the tread is lifting a hair if you get right down and study it.   The real test, however, became the Hallmark 4-8-4 with that fixed long wheelbase and all those scale flanges.  I had to tweak a couple curve on that one to run them out further raising the outer rail, wedging under the flex track, testing, and reballasting.   

[ednadolski]

I used .020 x .080" styrene strips cut into ~2" lengths, placed under the outer rail.  I transitioned into the curves with a few lengths of .010 and .015" stock for good measure.  It has worked very well, but it is very subtle and I think going as high as .030 would have looked more dramatic, and would probably have performed fine.

Happy Thanksgiving.

The speed limit on the Loop is ~25 mph or so, which IIRC would be consistent with a superelevation of about 2-3 inches.   So 0.020" max would be about right (allowing a few thou for license ).

I'm nor sure if the Walong siding is supposed to be superelevated, but it is built about 6" or so lower than the mainline.  I used some z-scale roadbed under mine, which is approx. 1/32" lower than typical N scale cork roadbed.  The effect is subtle, but noticeable on close observation.

Ed

[robert3985]

I use a variation of the masking tape method except I don't use strips that run the whole curve under the outside tie tips.  I mark the beginning of my curves in 1" increments (marks on the sides of my laminated Masonite subroadbed), then I number them from 0 to 5 on one end, then from 5 to 0 on the exit, 0 being the start of the curve on one end and the end of the curve on the other end.  in the middle of the inch increment marked "1", I place a single 1/2" long strip of blue masking tape that's about 1/8" wide, and then in the middle of the increment marked "2" I place a stack of two strips, continuing on each numbered increment until I hit "5" and then each consecutive increment gets a maximum five piece stack of blue masking tape until I get to the end of the curve and encounter the increment numbered "4"...which gets a stack that's four deep and so forth down until "0"...which doesn't get any tape. 

This does two things: (1) It gives me a smooth and consistent transition from flat to my maximum superelevation on either end of the curve, and (2) I "tack glue" my flex down in between my 1/2" long masking tape stacks/strips with gap-filling CA on the high side, and runny CA on the low side, using Accelerator.  This means the stability of my track's adhesion to my sanded cork roadbed is not dependent on the stability of masking tape adhesive over the years.

If the curve is short, I still mark it from beginning to end in 1" increments, but...if, for instance, there are only eight 1" increments, the highest elevation of the outside rail will only be three layers of masking tape, with two increments in the middle having those.

Since I use Micro Engineering code 55 flex, its stiffness allows me to get the curves nice and smooth before I start socking it down over the tape stacks.  I don't think this method would be as easy with Atlas "floppy" flex, but I could be wrong.  I've never done it with floppy flex so I can't say for sure.

For curves that are distinctly and prototypically highly supereleveated I've used as many as seven layers of tape.  My blue masking tape is around .009" thick. for a maximum of .063" of shimming under the outside edge of my ties on high-speed curves on my layout.

Here's what that looks like on my Echo Curve trackage, which is about 80% spiral easement with a short piece of constant curvature trackage in the middle of each eased entrance and exit.  I believe the minimum radius on the inside track is approximately 33":


Looks pretty steep, and part of that is because of foreshortening of the image because of the angle of the shot and the focal length of the lens, which emphasizes the superelevation.  I've run Big Boys, Challengers, MT-5's, GS-4's, FEF's, Mikes, Centennials, GTEL Superturbines etc., on this curve, including long container cars with body mounted couplers, and I've never had a problem, or stringlining with long trains up to 60 cars in length.

Just for giggles and as a comparison, here's the same curve on the prototype.  Superelevation looks about the same to me:


Here's the same curve from a different angle on my layout:


When running at shows or at home, the superelevation doesn't come right out and slap me across the face, but in photos, it REALLY shows up.

I think it's well worth the effort.

Additionally, I don't run any pizza cutters...every wheel has low-profile flanges on my layout, and that is not a hindrance on any of my superelevated curves...but then again, my absolute minimum radius is 24" so I'm sure those broad curves make a reliability difference on curves that are superelevated...especially with long wheelbase cars and engines.

[PGE_Modeller]

If I understand Bob's description correctly, his normal superelevation is 5 thicknesses of masking tape, total 0.045" (7 1/4 scale inches), between the inner end of the ties and a point 20 scale inches (1/8" actual) in from the outer end of the ties.  With Micro Engineering flex track, ties are approximately 9 scale feet long so the actual superelevation is 7 1/4" in 88" (108" - 20") which works out to be a 4 5/8" elevation difference between the inner and outer gauge points of the track.

There is also a difference in the way we do superelevation on our model railways and the way the prototype railways did it (at least in the steam era).  We keep the inner end of the ties at a constant elevation (in the absence of grades) while the prototype kept the gauge point of the inner rail at a constant elevation.

Cheers,

[Big Train]

Just a few cautions about super-elevating track to prototype standards. If too severe the ability to run longer trains with short wheel base rolling stack can cause the train to tip inward and fall off the rails in some situations.

The other thing we noticed is that some rolling stock, unless adjusted correctly, that is the bolster pin or screw adjusted to prevent excess slop, may result in the rolling stock flopping over excessively, more than the expected result that was intended. This was particularly seen with earlier version of the Rapido (sorry Puddington...) passenger coaches.  In some cases I removed the interior lighting kits to lower the centre of gravity to aid in preventing the wobble.

When I added super-elevation to my modules, it was just enough to give the effect without compromising train operability.

Hope this helps