Superelevation

[TheFlyingScotsman]

I am toying with  the idea of a layout rebuild and while I am far from being an expert modler in the sense that some around here are but I do know how to lay billiard table flat trackwork which i find is the cornerstone of an anxiety free running session. 

So my question is this - I am tempted to superelevate my curves (min rad 40") for the look but for those who have experience of this does it mean less reliable running?

[NHTX]

     If you run long trains, it can increase the likelihood of string lining derailments.

[7j43k]

I've done it, though on 60" curves, where not much is likely to go wrong, anyway.

I did about .025" of superelevation of the rails, which isn't very much.  That amount shows, but it isn't OBVIOUS.  Subtle.

You should also use easements.  Then you introduce the superelevation in that easement.

I doubt you'll have trouble with 40" curves and "minimal" superelevation, combined with those easements.

Ed

[tstage]

On my previous layout I used 0.03" superelevation on a 225^o section of R22" curves, with easements on both ends - including the superelevation.  It was noticeable enough and I didn't experience any derailments that I recollect.

Tom

[selector]

"... I am tempted to superelevate my curves (min rad 40") for the look but for those who have experience of this does it mean less reliable running?"

I can.  Generally, I do very well with this, but occasionally, often well into my testing, one locomotive will come off the rails somewhere along a supered curve. It can be a bear trying to figure out what is going wrong, but it generally means the height disparity at some point is too much for whichever axle/truck is lifting out of the gauge.

So, those of us who chime into threads on this topic tend to caution people attempting it.  Less is more in this case.  In the real world, depending on the corridor and speeds, super is going to be between 1 and six inches across the two rails.  That's pretty slim in our scale world, approximately, about 5 thou for a 4" cant on the outer rail.

If you try this, ramp up to the apex gradually, and DO NOT sand or plane the inner rail's roadbed such that it also dips. Only fashion a ramp up and then back down for the outer rail.

 

Dick-et-al-2016-AREMA.pdf (illinois.edu)

 

 

["JaBear"]

At the Club, I raised the idea of having superelevation on the mainline curves, 24” min. radius, and was told a resounding NO! The reason being that it would cause “string lining” derailments. However as, at that time, I was basically the only one building the benchwork and tracklaying, I read further on the subject and decided to give it a go. If it was a failure, then I was the only one to have wasted time.

 

So, taking my time, on a 90° 24” radius curve using easements I ended up with a maximum superelevation of 0.30”, which then translated into a long “S” curve which I also super elevated to 0.030”. There was also a slight grade change as well.

 

Even though some of the Club members had rolling stock whose rolling characteristics were dubious to say the least, that section of track never gave any trouble.

 

The sad thing is that while, for those who “looked” a long train looked “cool” going through that section, a lot of the members never realised that I had built in the super elevation!

 

It’s not a technical term but I strive to achieve “flowing” trackwork.

 

Cheers, the Bear.

[gmpullman]

It’s not a technical term but I strive to achieve “flowing” trackwork.

I'm in Bear's camp (Den?) on this as well.

PRR_T-E7_sm by Edmund, on Flickr

I built my layout in '95 and wanted to try superelevated curves. I'm using Shinohara code 83 marketed, at the time, by Walthers. This is ideal track for applying shims to the outer rail as the spikes are located along the rails, not in the tie centers.

I do not recall if I used strips of .030 or .040 Evergreen styrene for the maximum elevation. I did use a gradual "stepped" incline with varying strips at .005" intervals beginning at the easements into the curves. The transitions were about 14" long for each step as they entered the curves, that is, flat = tangent, then .005, .010, .015 — etc. until reaching (.030 or .040) for the majority of the curve.

Even on one of those controversial reverse curves, hasn't been a problem.

IMG_4576_fix by Edmund, on Flickr

Stringlining? Phooey. Every once in a while I've run ninety-car freights. Sometimes eighteen-car passenger trains. Never a problem. There is the occasional piece of rolling stock that will refuse to abdicate but it is a fault of the car, not the track.

I've recently had to modify many of my Broadway PRR P70s because they use a truck bolster that is just too snug. Otherwise, no problems.

Never regretted laying track with superelevations.

NYC_1646_F7 by Edmund, on Flickr

Good Luck, Ed

 

[riogrande5761]

Ed, looks like you superelevated the camera in that first photo. 

[gmpullman]

riogrande5761

Ed, looks like you superelevated the camera in that first photo.

Here's a shot from another time. Note the caboose leaning port and the distant box car leaning starboard:

BnO_4011-FA1 by Edmund, on Flickr

Maybe it is because in the other shot there is an angled retaining wall where the track goes up a 2% grade? Here is a tighter version of a similar photo.

PRR_T-E1_sm by Edmund, on Flickr

The superelevation is subtle. Only the pair of main line tracks are superelevated. The outer tracks are flat.

I'll place a digital level across the rails later tonight.

Regards, Ed

[SeeYou190]

I design/build my layouts with nearly all the curved trackage hidden.

On my previous layout, I had a tunnel portal on a curve. The curve through this portal was 36 inch radius. I superelevated the track at the portal only for photograph staging purposes.

It looked amazing, but as far as I know, I do not have any pictures of the "spare bedroom" layout, sorry.

There were never any operational problems.

-Kevin

[7j43k]

There are two ways modelers can describe the extent of superelevation.  One is the difference in height of the two rails (the official way).  The other is the difference in height of each end of the ties (the amount the ties are shimmed).

Different numbers.

While my rails have a superelevation of .025", my ties have a superelevation of .040".  That reflects the thickness of the shims under the outer edge of the ties.

Since we are trying to be helpful by exchanging information, accurate description of what we've done is important.  I recommend using the Official Way.  That number can of course be augmented by describing how you got there.  Like using .040" shims under the ties.

Ed makes a good point about placing the shims under the outer rail, instead of under the end of the tie.  My view was, and is, that once the ballast is applied, it will do a fine job of keeping the superelevation.  So I shimmed under the tie ends.  My spiking (into cork) was done rather gently.

IF I had done my shimming as Ed described, my superelevation would have been higher, by the way:  .028".  Not lots, of course, about 10%.

 

Ed

[Pruitt]

I use 1/4" masking tape to superelevate my curves. On this first one I ever did, I used 1/2" and it buckled along the inside of the tape. The tape is centered under the outside rail. Different brands of tape vary in thickness. Mine was about 1.7 thousandths of an inch. This was on a 30" radius in HO:

That's five layers of masking tape with each layer staggered from the previous one by about 1 1/4". Since then I've increase to 6-7 layers. This gives roughly a scale 1 1/2" elevation of the ourtside rail over the inner.

How the train looked once the track was laid:

I've always superelevated curves since then. 

Superelevation can encourage stringlining, but not by much unless your superelevation is excessive. I've pulled 40 car freights around my 30" radius curves with no issues whatsoever.

[selector]

On spline roadbed, I plane the roadbed so that it provides super:

On strip roadbed, including cork and foam, I use strips of the clear plastic packaging you get from so many durable items hanging in rows at the big box stores: [It's subtle, but discernible...]

[doctorwayne]

When I created the 45' long and curving grade from the layout's main level to the partial upper level, I first fixed the bottom and top on 90º risers at appropriate heights.  I then measured to the mid-point of the grade and placed another riser, also perpendicular at 90º, and at a height one half of the total grade.  I then continued adding risers, more-or-less dictated by the height of the 3/4" thick plywood sub-roadbed.  At this point, all of the risers were simply clamped to the open grid 1"x4"s, with a pencil line on each, even with the top of the grid members.

Next, I built a train on one of the curved segments of the sub-roadbed, then removed three of the clamps at roughly mid-curve.  I then deflected the bottom of the middle riser until the tilt of the train looked reasonable to me.  I then re-clamped that riser in place, with the inner end of the pencil line even with the top of the grid member.  Obviously, the outer end of that line was somewhat higher.

Next, I unclamped all of the risers within that curve, allowing their bottom ends to deflect outwards to diminishing degrees as they moved away from the mid-point.  I then, without altering the deflection, raised each, in-turn, so that the inner end of the pencil line was again even with the top of the grid member, the outer end of the lines diminishing in height as they moved away from the mid-point.

Each of the five curves on this grade was superelevated in the same manner, which automatically introduced vertical easements in-to and out-of the superelevation. (The curves already included horizontal easements.)

Until today, I had never measured the actual superelevation...it looked reasonable and never caused string-lining, even on reasonably long trains.

I just finished checking with a level and some various sizes of strip styrene, and pretty-well all of the four on-ground curves have a mid-point superelevation of .040", while the one that's mostly on a bridge (and a curve of more than 180º) has a superelevation of .030".

Creating superelevation in this manner was really easy, although I'm sure that anybody using this method would have their own opinion on what degree of superelevation looks good to them.

Wayne

 

["JaBear"]

7j43k

There are two ways modelers can describe the extent of superelevation.  One is the difference in height of the two rails (the official way).  The other is the difference in height of each end of the ties (the amount the ties are shimmed). Different numbers.

Ok Ed, that’s a good point you’ve raised, so following from my original reply, for clarities sake, I shimmed in increments, using 0.005” card the outside rail to a total shim thickness of 0.030”.

 

Cheers, the Bear.

[TheFlyingScotsman]

Thank you everybody for the responses and those beautiful pictures. There's a lot to digest here but I think come the time I'll lay a section and run all the rolling stock to get a feel for how things run.