I’ve seen term “reverse curve.” What does that mean?
I believe it’s the same thing as an S-curve. Where it starts going in one direction and then goes in the opposite.
PS–I just typed “railroad reverse curve” into Google and that’s what it is.
IF you don’t have several hundred feet between those curves turning in opposite directions, you have a potential nightmare on your hands.
I new you’d go off on a tangent, there.[:-^]
You know MC - always on the straight and narrow…
The sharper the curve in degrees the greater the distance needed. Also, the speed which trains operate is involved.
BNSF (former Santa Fe) constructed a new line through the mountains of Arizona in 1959-1960 which allowed passenger trains to operate at 90 MPH. and had one degree (or less) curves. The minimum distance between curves was 500 feet of tangent.
Would one of the reasons for needing 500 feet of tangent be dealing with transitioning superelevation? For model railroads where track isn’t canted, allowing for the length of the longest car should be enough.
Superelevation follows the “spiral easement” entering or exiting a curve. The tangent between curve allows the center lines of equipment to line up straight before starting a new curve.
rcdrye, you are correct.
I believe Santa Fe used 570 foot spirals with a gradual step up to three and a half inches of super elevation. But the 500 feet of intervening tangent still applied.
Thank you, gents.
That’s pretty much what I thought could be the reason for the 500’ tangent. Since the purpose of superelevation was to balance out “centrifugal force”, the amount of superelevation would increase with the increase in curvature of the spiral.
(1) Before you get into the superelevation, you have an unstable situation with trucks turned in opposite directions.
(2) Track liners, given enough time and iterations, will overlap the spirals.
(3) Speed will have to drop mui-pronto or the gods of centripital force will induce some severe rock and roll. (think wheel lift) Cars and locomotives leaning in all directions. Sometimes you have no choice and train handling gets iffy. (time to get absolution from the boss and the Chief Engineer)
(4) If anything in the consist has centerbound or stuck trucks - GAME OVER
(5)and yes, your superelevation will accentuate the rock and roll.
(6) having seen what happens on Raton, Glorietta, Tennessee, Moffat and Cajon, grade and braking only makes the issues worse when added to reverse curves. (bad enough in some yards)
mudchicken" - Track liners, given enough time and iterations, will overlap the spirals.
MC, with his many years of latest experience, knows that the original engineering construction design disappears with changes in train operation and maintenance procedures. I probably would not recognize the engineering I participated in creating in 1959-1960 if I were revisit Williams-Crookton, Arizona.
Klauder taught me that the reason for the ‘tangent’ between spirals was dynamic; the resultant of curve and superelevation relief rocks the car on its springs and damping in roll is relatively poor. It can be worse if the truck roll couples to yaw and there is any slop between the bolster and sideframes.
Makes sense, as traversing into or out of the spiral requires changes in angular momentum along the long axis of the car (ie. changes in rate of roll). I would perhaps naively assume that easing into a spiral from a tangent with the corresponding easing into changes in superelevation as well easing out of the spiral’s increase in curvature to the constant curvature portion of the curve would reduce some of these dynamic effects. In flying terms, this would be equivalent to smoothly move the yoke or stick in changing the bank angle as opposed to a sudden movement of the yoke/stick.
Similar thing with straighline acceleration - a perfectly rigid body would not have any dynamic effects of “jerk”, were as non-rigid body would show sigficant differences between a given acceleration applied instantly versus applied gradualy.
Fun to see a lot of interesting info in this thread.
One only need observe a Lego train to appreciate the spiral…
Or a model train running on snap-track.
I’ll be interested to see the rules diningcar used in incrementing superelevation (which is also entry- and exit-spiraled) in his 570’ curve spiral. My somewhat naive preference is to commence the superelevation after the car has settled into the spiral in yaw, and then have it fully relieved before the lateral spiral goes back to tangent. A major point of superelevation is to decrease overturning load on the low rail, so both speed and degree of curvature factor into how much is used – this is different from ‘passenger’ use of superelevation which I’d expect to be coordinated with actual lateral angular acceleration in the spiral.
I shall try to remember; Santa Fe had a standard plan for spirals that was in a printed format which each Party Chief carried.
The curves were staked (in those days) every fifty feet as the 570 foot spiral was created and deflections for the transit operator were calculated to those points. The 570 spiral point was established from which a one degree curve was the established every fifty feet until the spiral at the other curve end was created back toward the (at least) five hundred feet to tangent.
Superelevation was calculated to coincide with the fifty foot centerline locations and was increased incrementally to the 3-1/2 max, which then continued until the reversing spiral was reached
You prefer to have no superelevation on what, the first 100 feet of the spiral? Does any US railroad do that?
For a given speed, the amount of superelevation in a curve should be based on the radius of that curve. The tighter the curve, the greater the superelevation.
If the radius is infinite, no superelevation is needed. If the radius is “huge”, very little is needed. If the radius is “sharp”, much more is needed.
At any point in the easement for a curve, there is a particular radius. And there is an appropriate superelevation. It starts from zero, because of the (near) infinite initial curve radius, and increases as the radius gets tighter, reaching its maximum in the “true curve”.
Thus, all changes in superelevation should happen in the easement of the curve.
If you’re going to do something different, you better have a VERY good reason. Do you?
Ed