Someone posted a question about “track banking” on one of the other fourms recently. That got me to thinking:
Thanks!
Tom
The center of gravity is somewhere above the wheel centers depending on the weight and amount of water in the boiler. Depending on the speed and height there are outward forces on the center of gravity that will tend to tip the locomotive with the flange against the outer rail. In the nineteenth century as locomotive power and speeds increased it was found that superelevating the outer rail helped to counter this tipping force. A lot depends on the tightness of the curve and the speed. I would not bother superelevating anything under 24" radius and will add that on my 72" radius I use 1/16" under the outer edge of the ties. This probably reflects the 4" to 6" difference in rail height of the prototype. It is a beautiful thing to watch a freight train take the curve at 35-40 mph, or even higher for my passenger train, say 50-60 mph. It doesn’t take much and is easy to overdo. I hope this helps your decision. jc5729 John Colley, Port Townsend, WA
I know that superelevating has been around for a long time, but your question got me to wondering more.
I haven’t found an exact answer, but I found this great site that has all kinds of info on it. Under the heading of “Railway Civil Engineering, Historical and Current” is a link to a page called “Transition Spiral”. In the first couple of paragraphs it mentions superelevations and gives a hint that this might have been a very early development in railway design. It kind of sounds like it might be pre civil war.
I’ll keep looking and watching this thread, because now my curiosity is peaked.
John,
Thanks for the detailed explanation on superelevation. I already under the reasons for it. I was really only looking for historical information as to when it came into being, not why.
Tom
Number 1 has been answered.
Number 2; The first RR in America designed for steam locomotives and not horses was the P&R later the Reading - between the city of Reading PA and Philadelphia (Today, it is still an actively used Norfolk/Southern line).
3 miles below Reading PA is the Klapperthal curve; a graceful sweeping curve along the Schuykill River that featured super elevated curves. It was built in the 1830’s and its surveyor & chief engineer, Moncure Robinson from Virginia cut no corners. I have a photo from the 1870’s showing the track banked and superelevated.
Similar photos of the PRR’s Horseshoe curve near Altoona PA, from the late 1850’s or 60’s shows that curve as being superelevated. It was originally 2 tracks and later enlarged.
Super elevating track depends on the speed, load and inertia as well as the degree of curviture, change of grade and the transition to & from the curve. Also banking & anchoring the track at a curve actually helps preserve the alignment of the track from kinks or buckling. When a train exceeds 20 mph through a curve, load shifting, changes in the ‘center of gravity,’ tipping and other physics take over. There’s alot of civil engineering going on here.
When train speeds increased to 40, 50, 60 to even 100 mph at the turn of the last century, the individual railroads rebuilt and re-engineered their track so they could run faster and heavier trains.
Until 1970 I believe, the individual roads regulated their track speeds and conditions until the Federal Railroad Administration took over setting the track standards and running conditions.
See this item; http://www.zetatech.com/CORPQIII45.html
It discusses the problems and costs of operating passenger trains on roads having slower freight traffic with respect to track design and conditions.
At any rate, there sure is a lot of beauty to any train when it leans into a curve at speed.
</
Mark300,
ICC speed limits (79mph passenger, 69mph freight) were mandated in 1951. Trains equipped with cab signals/automatic train stop were permitted to exceed those limits.
Individual railroads continued to set their own maximum speed and identify their own restricted speed zones, based on traffic, rail geometry, locomotive capability and local geography, and still do. Quite frequently, the company-mandated speed limit is considerably lower than the Federal standard.
As was noted above, superelevation was begun early - just about the time that the people aboard the earliest trains noticed, and were alarmed by, the side thrust generated by lack of superelevation. For modeling purposes, superelevation is purely cosmetic - but it definitely increases the realism of the modeled scene.
Chuck (modeling Central Japan in September, 1964 - with superelevated curves and spiral easements)
a. No block system in effect:*
49 mph freight, 59 mph passenger
b. Block system in effect:
79 mph freight, 79 mph passenger
c. Automatic Train Stop or system meeting 49 CFR 236 Subpart H installed:
Maximum speed determined by characteristics of system.
Speed limits on curves are mandated by the FRA according to the amount of superelevation in the curve (49 CFR 213.57), the degree of curvature, and the class of track. The FRA allows up to 4" of unbalance but most railroads set a maximum of 3" unbalance; at least one Class I now only allows 1.5" unbalance. (Unbalance is the difference between the mathematically determined superelevation required to balance the centrifugal force for the speed of the train, and the actual amount of superelevation, which is generally less than the mathematically determined number. The civil engineer, knowing the superelevation that is present in a curve of X degrees, adds to that number the unbalance that the railroad will permit, up to but not exceeding the FRA maximum, and consults a look-up table to determine the maximum speed thus allowable in the curve. Or, the engineer can start with the desired speed, consult the table for that degree of curvature, and determine the superelevation that is required, less the maximum permitted unbalance.)
Superelevation was introduced prior to the 1850s in Britain. It probably entered about the same time in the U.S. but did not become very common until the 1870s as train speeds
I’d have to search for written documentation of it, but I’d pretty much guarantee that I.K. Brunel’s GWR was built with superelevation back in the 1840’s-50’s when track was first being laid.
Not to re-start an old discussion, but although superelevation isn’t really needed for model trains re speed and such, it does actually slightly lessen the effect of the curvature - i.e. makes the curve slightly less sharp for the train, though only by a minor degree - so it isn’t purely cosmetic on a model layout.
This makes sense to me. With the introduction of the American Standard in 1840 trains would be moving faster. Easements would be developed to keep the flanges from ‘clunking’ at the beginning of a turn by taking up the gap between the flange and rail before the actual turn started.
Since flange drag sucks power, it would make sense to bank the curve a bit to keep the flanges from riding that outside rail all the way through the turn.
These forces would have been present and noticable early on, and be refined to a fine art as quickly as the controllers (beancounters) became aware of the economic consequences.
Modernly, I hear flanges only touch the rails at a switch where easements and superelevation isn’t practical.
Or, I could be mistaken,
-rrick
I have an English civil engineering text, published in 1867, which describes methods for determining superelevation as “having been developed and perfected these many years hence”. That suggests to me that by 1867, superelevation had already been around for quite a while.
A classic old Model Railroader cartoon showed a guy who had superelevated the wrong side of the rail on a curve, with the engine laying on its side … almost ranking with the cartoon of the guy with benchwork that was at all sorts of crazy elevations. He is holding his level and saying “Well I’ll be darned the little bubble was stuck …”
Dave Nelson
Prototype railroads almost all had superelevated curves and spiral easements, ergo model railroads should as well. The amount is a matter of what appeals to your own eye. Reducing the prototype dimension to scale is meaningless, it is soo small. Do what looks good to you and You will never go back to running a toy train set again.