How do they do that?

[Willy2]

Simple question to ask. Most likely, hard question to answer.

How do trains manage to stay on the tracks as much as they do? I'm still amazed that all of that weight can stay on the thin rails at 70 mph.

Willy

[Modelcar]

Willy.....For many years I've tried to understand "how" to the question you asked...I have tried to understand the strength of the rail under the train and still it seems to be a question in my mind how the rail stands such forces. When one consumes the image of one of our larger engines standing and then eyeballing the rail it is standing on...there seems to be no question, the rail should simply be forced to turn on it's side with such weight...and thinking beyond to the forces that come into play as that engine enters a curve and that then adds side forces to the track structure....?? But it seems to do the job...I've watched trains move over rail so out of shape and wavy and sometimes being light rail...on lightly used branches and wonder how it travels 10 ft. without going on the ground....But for the most part they stay where they're supposed to...One example: Track on Florida Central around Eustus and Tavares is horrible but it remains useable....

[dehusman]

Gravity and friction.

[Anonymous]

Sorry to do this, but I can't resist. I'm going to give you something else to think about. I've had many experiences where I will come to a grade crosssing and there will be a train coming around 50. I look closely, and about three feet off the crossing, they fly about two inches above the rails for about a foot, and that's including flanges. Think about that, because I certainly have been!!!!!!!!!!!!

[Anonymous]

I have always been intrigued by double stacked containers which don't tip over because of their high center of gravity. It always seems to me they will roll over on a very sharp curve.
Hmmm!

[Anonymous]

Don't they put the heavier containers on bottom to balnce the weight out?

[garr]

Willy,

My background is not physics or engineering so I imagine you will get a better answer later. The two major factors at work are gravity and centrifigal force.

The flanges along with the weight and gravity do a great job keeping the train on the tracks. The rail top is actually a lot wider than the area a properly contoured wheel on a properly contoured rail needs to ride safely(Think of a marble on a table top). The contact point of rail and wheel is very small in relation to the size of the train. Someone in the engineering department of a railroad once told me if everything is in proper contour the square inches of wheel/rail contact on a typical coal train is not much larger than a legal size sheet of paper. As long as the flange is not worn, such as to a point, it can not "bite" into the rail to ride up. Remember there is very little friction between steel rail and steel wheel.

If the trains are traveling at 70 mph the train overcomes the centrifigal force on curves with (super)elevation. Even curves built for much lower speeds will have elevation built into them to fight this force. The outside rail will always be higher than the inside rail on these type of curves.

When something goes wrong between/with these elements plus misaligned, out-of-gauge, or improperly maintained track is when things start falling apart or off the track as it may be.

Jay

[Modelcar]

...I've noticed at many crossings the track structure is low at each end of the street crossing and the rails travel up and down several inches in some cases and slam down as each truck passes over it...etc...
And has anyone ever wondered about the heavy timbers just inside each rail at a street crossing to form an inside edge for the flange to travel and what risk it might be for one of those to just get moved out of place a bit that would contact the flange or wheel itself....but there again, haven't heard of much problem in that area either.

[Anonymous]

I think the low track structure on either side of the street is the reason you here the classic "duh-duh..duh-duh.....................duh-duh..duh-duh."

[garr]

Modelcar,

Again I think the weight would handle the misaligned timber if the track is still in gauge. Ties are usually beaten badly when a car or two are derailed mid train and the train continues(until noticed or the air line disconnects or the cars topple). The locomotive would be the first to hit the timber so it seems the timber should lose in that contest.

Jay

[Anonymous]

QUOTE: Originally posted by joantvw

Don't they put the heavier containers on bottom to balnce the weight out?

Can't argue with you, however if there are 2 containers of a different length, from my observation, the longer one in 90% of the cases is on top. I don't know if
weight is a consideration.

Anybody else have any answers?

[tree68]

Some years ago I was stationed in Rantoul, IL, and frequently did some trainwatching on the platform of the ICG station there. The track was less that perfect, and the locos would be swaying pretty good as they came by the platform. I sometimes wondered if I should be looking for an out...

[Modelcar]

Yea, Jay...weight may be what saves many trains from derailing under some of the circumstances we've been discussing....Sometimes these rail facts just seem to defy reality though. I guess what I am saying in general....It seems to me the mechanics of it all should put more trains on the ground but for some reason it all seems to hang together and work every day as a routine and for the most part, keeps on working.

[Anonymous]

Another reason it stays on the track (except on the UP in San Antonio [banghead][banghead]) is that the wheels themselves are coned--that is, the profile of the wheel tread slopes upward from the flange to the outer edge at the railhead, so that the diameter of the tread at the flange is larger than the diameter at the outside edge. This slope of the wheel treads, when riding on the railheads, (HEY, I'M A POET!!) tries to force the flanges to center themselves between the rails. When one wheel tries to crawl up the railhead toward the flange, the coning on that wheel tries to push it back and the coning on the opposite wheel tries to pull it back.

It doesn't collapse the rail (usually!) even with the very thin web, because all of the strength in the beam (which is what the rail is) is aligned vertically in the web. However, if the rail base is not affixed well to the tie (say, if the tie is rotten or spike-killed, where the spike holes have grown too large and the spikes won't hold well), any substantial weight and/or lateral force can roll the rail right over. Even with a standing car--a bad habit that was identified with the Deramus-era Katy in the 1960's, and some other roads as well.

I have been amazed at what a set of rails - even old ones - will hold up. On some of the properties we have rehabbed, as a practical matter, the mud and the weeds were holding the rails up in spots, and you could still get a train over it (but not very fast). We took out some 61-1/2 lb stuff rolled in the 1880s that was still sort of in occasional service on an old branch line, and would still hold up for a 210,000 lb weight limit (70 ton car).

[Anonymous]

QUOTE: I have always been intrigued by double stacked containers which don't tip over because of their high center of gravity. It always seems to me they will roll over on a very sharp curve.

And I think this is the best reason to limit the narrow gauge trains running in several countries including mine to operate double stacks even if it is economically feasible to do so.

Karn[:)]

[Modelcar]

Drephpe...Yes, what you have related in the questionable old track and still able to support the weight of a train is simply amazing. I note even on main line track you can see spikes sticking up above the rail base.

[Willy2]

Thank you for the answers. It seems to be a complicated thing. But I think I'm beginning to understand it all. [:o)]

Willy

[Hugh Jampton]

It's always great that the simplest of questions have the most complicated answers. Plus, to describe this I would normally use a pencil and a bit of paper and do some drawings, so this may seem a bit long winded.

There are 2 forces involved, the vertical force (proportional to the weight and gravity).
Taking the vertical forces first; a 286,000 lb gross car on 4 axles gives a static load per axle (assume 2 2 axle trucks) of 35.75 tons, or 17.875 tons per wheel. There's 2 ways in which steel can fail, shear and bending. Shear first. 136 lb rail has a cross sectional area of 13.30 square inches, so the shear load is 1.34 tons per square inch. Steel can carry about 12 tons per square inch in shear,, depending on what type it is (yes, there's more than 1 kind of steel). This gives a factor of safety of 9. Sher failure will happen usually above the tie because that's where the shear force is greatest. Bending is greatest when the load is applied mid-way between adjacent ties. Bending failure has to do with the shape of the rail and the distance to the support (As an experiment take a plastic ruler and lay it flat on a table with one end hanging over the edge. Apply a load, and see how much the ruler bends. Now turn it on its side and apply the same load. The ruler bend much less. This is called the moment of inertia) 136 lb rail has a moment of inertia of 93.94 inches to the 4th power (don't worry about this, it's the number that's important). So our 17.875 wheel load will create a bending stress in the foot of the rail (where bending failures happen) of 7.9 tons per square inch. Rail steel can carry a maximum bending stress of 50 tons per square inch. So this gives a factor of safety of 6. One thing though,, these figures are based on the static load (the car just sitting there) Dynamic forces can double the stresses in the rail, and impact forces from wheel flats etc can increase the force even further.

There are 2 main lateral forces which need to be considered. The first of these is the lateral force that always occurs and tries to turn the rail onto its side. As mentioned before, wheels are conical in shape (the radius at the flange {gauge side} is greater than the radius at the bearing {field} side. The downward force on the foot of the rail provided by the spikes is multiplied by the horizontal distance to the centerline of the rail. This is compared with the lateral force of the wheel on the rail multiplied by the height of the rail. (these multiplications are called moments). So as long as the moment at the fixing is greater than the moment at the wheel rail interface the rail will remain upright. The moment at the wheel rail interface is reduced by inclining the rail inward at the same angle as the conicity of the wheel. On curves, where the lateral force also includes curving forces the rail is usually held at each tie by 2 spikes, and the outer rail of the curve is raised to give some superelevation.

Wait,, there's more. Most people seem to be under the impression that the flanges guide the train around a curve. This is of course false in most cases. The train is guided around the curve by the conicity of the wheels. Because the 2 wheels of an axle are held rigidly together by the axle there can be no relative movement between them. So in a curve the axle moves off centre to the point where the ratio of the 2 different wheel diameters equals the ratio of the 2 different curve radii of the rails (the radius of a curve is measured to the track centerline. The outside radius equals the curve radius plus half the gauge and the inside radius equals the curve radius minus half the gauge). Trains don't tip over in curves (sometimes they do) because the superelevation reduces the centrifugal force. But for the train to tip over the centrifugal force (which is the speed times the speed times the radius of the curve) times the height of the center of gravity above the rails has to be grater than the wheel load on the inside of the curve times half the gauge. As an example; a 286,000 lb car with a center of gravity 9 feet above the rail travelling around a 500 foot radius curve (with no superelevation to make the math simpler) must travel at greater than 80 miles per hour for it to tip over.

Sorry if it's a bit arithmaticy, but that's the way it is sometimes, but if you sit down with a pencil and the back of an old envelope and draw it all out I'm sure it will make some semblance of sense.

[Modelcar]

Hugh...I've read all of your data carefully. I doubt none of it. I like the data used to describe the wheels following around a curve. Conical shape of the wheel treads have to be as designed to do all this in a neutral manner. After wear sets in then I suppose it tracks a bit differently. Of course rail head shape has to be as designed too, to help all this neutral work take place. Really complicated and interesting stuff...

[Anonymous]

Hugh--

Nice work. I was afraid to throw the math out there, but I think your explanation can be understood easily by the readers. One note, though, unfortunately in most of the US the flanges by default do wind up guiding the axles around a lot of curves, particularly the sharp ones in heavy-haul land. This is exacerbated by the fact that, as a general rule, the US does not superelevate to balanced (ideal) conditions, primarily due to large discrepancies in train speeds on a given stretch of track. What this means is that we see a huge amount of curve wear on the gauge side of the outside railhead, accompanied by matching flange wear. This is especially evident on large heavy cars, and not just on higher speed lines. Of course, truck hunting doesn't help, either.

Modelcar--

Yeah, that 61-1/2 lb stuff was amazing. It's truly antique, and to find 120 year old rail still in service on a main track is obviously very, very unusual. I think we got a small sample when it was pulled out and replaced with some really nice 90 lb relay. You may ask, why 90 lb, when weight limits are effectively obsoleting it? And the answer, like in all other situations, was $$. Weight limit for that stretch of line was set by a couple of bridges that the client could not afford to upgrade, so the line stayed rated at 210,000. For very light traffic and no heavy grain or aggregate cars, the result was far more serviceable than what we started with.

Happy fourth from the only recognized independent nation to join up with the colonies![:D][C):-)][bday]

[Modelcar]

drephpe....I'm sure the 90 # rail was a great improvement over the 61 1/ 2 stuff.....120 years old and in service yet....!! That is pretty wild.

[dehusman]

Nobody has mentioned that the rails themselves are canted (typically 40 to 1 ) inward to match the taper on the wheels. The cant is built into the tie plates.

Dave H.

[Anonymous]

QUOTE: Originally posted by dehusman

Nobody has mentioned that the rails themselves are canted (typically 40 to 1 ) inward to match the taper on the wheels. The cant is built into the tie plates.

Dave H.

True. That assumes, by the way, that there are tie plates there and that they work. A safe assumption on US mainlines and heavily-used branches, but not necessarily the case elsewhere in the world (or some situations in the US).

But we should have mentioned them anyway. Thanks.[oops]