I have for awhile carried this thought off to the side pondering whether or not it mattered which side of the wheel the flanges were on. What if every wheel set had them over the outer rail edge rather than inner? Diamonds and grade crossings could be properly negotiated with some modifications, but what about switches? Is it possible to design a switch accomodating exterior flanges?
Switches might be more complicated, with a different arrangement for the points and frogs. Anything can be done, but what would be the point?
I’m willing to bet that it was one of those ideas that was contemplated back in the beginning, and was discarded in favor of the arrangement we now have.
Old Timer
What I’m thinking is with exterior flanges, there would be an outward pull on wheels, especially in curves, creating more stress on certain parts, whereas, on the inside there is a natural tendency of push inward (less stress on the wheel to axle connections).
Right you are, MWH.
I didn’t even look to see who posed the question. I should have known.
Next time, I will.
Old Timer
No they’re not! Give it a second thought and envision all the physical forces at play. It’s quite easily conceived. Aside from this however, you are right on all other counts throughout this brief discussion.
Many many years ago, this question came up in (I think) Model Railroader (please don’t make me try to look it up)! Their explanation was that the vehicle would try to move to the OUTSIDE of the curve, and that the wheel tread was tapered, rather than flat across, therefore the wheel on the inside would be operating on a smaller diameter, which was desirable since the inside wheel traveled a shorter distance and there was no differential like on an automobile. Otherwise one wheel or the other would be slipping or dragging around the curve.
Don’t shoot me; that was THEIR story!
(This explanation was prompted by someone asking how rail vehicles got away without having a differential in each axle).
I’ll try to find it, Mark; it was probably back in the '60’s or early '70’s so might take a while. I hadn’t really thought about it. The explanation made sense so I just accepted it!
Two different physical effects being a bit confused here:
Curve following without draft-force effect: the vehicle tends to the “outside” of the curve via the mechanism that produces “centrifugal force” – thus eventually loading the outer-wheel flanges (if inside the gauge) or the inner-wheel flange (if outside).
Draft force, as mentioned, ‘stringlines’ the consist, leading to resultant force directed inward as described.
But note that one of the purposes of coned tread and matching railhead taper is to provide restoring force for tracking and steering without actual flange contact. If you model the geometry here, you get divergent, not convergent geometry with outside flanges (if the taper blends into the outside flange, like a conventional railroad wheel turned around) and a nasty stress raiser if you have a conventional conical tread profile that then reverses at the ‘gauge corner’.
One point I believe posters were trying to establish is that a reverse-taper tread won’t produce the ‘differential’ effect when going around curves – expect more wear, noise, vibration, rail corrugation and martensite cracking, etc. etc. etc.
The ‘force on the wheel-rail connection’ is just about exactly reversed when you reverse the position of the wheel on the axle, ceteris paribus. Force on the outside journals, bending moment in the axle, etc. are not the same – but that wasn’t the question. The effective incident angle of perturbing force, the geometry, and the wheel and axle size and material may be taken as common, and at least in theory everything (in this part of the discussion) is rigid. There isn’t any suspension or compensation between wheel and axle, and no damping between flange and railhead. If you were to factor in the characteristics of primary and secondary suspension, roll, truck skew, etc. there might be some induced force components that are a bit different, but I wouldn’t write home to Mother about their magnitude… ;-}
Asid
You might want to check out U.S. Patent 3,707,125 at
http://v3.espacenet.com/origdoc?DB=EPODOC&IDX=US3707125&F=0&QPN=US3707125
(if the link is garbled, check for U.S. Patent 3,707,125 at ep.espacenet.com)
This steering arrangement as for the GATX RRollway (the GATX of tank car fame). The plan was for a wide-gauge railroad to transport passengers inside their cars where the cars are driven in from the side – the rail cars are 20 feet wide. I saw a model of this as a child where they ran the thing on pairs of HO gauge tracks – I guess the model builder didn’t want to hand lay track for the wide gauge. The Chicago Tribune didn’t know any better and showed an “artist’s concept” drawing of the thing running on a pair of standard gauge tracks – that was never the plan.
If you look at the drawings, they had a pair of sideways rollers on only one rail – the wheels on the opposite rails are steered through rods. Mechanical engineers call this a “four-bar parallelogram linkage.” This system would have needed a special kind of frogless switch.
To show you how much times have changed, note how the cars are parked on the train in pairs – there are crosswise passage ways to get in and out of the cars only between pairs of cars. A friend pointed out that on modern cars, either the driver or the passenger would have to crawl over the center console-gear shift – common even on auto transmission cars these days.
There is one operator that uses outside flanges. Israel’s currently ONLY subway (with one being planned for Tel Aviv), the “Carmelit” because it runs up Mt. Carmel (the Biblical Mt. Carmel) in built up areas of Haifa. It is completely underground, single track with one centered passing siding, six one-platform stations, three above and three below the siding, symetrically placed, and is a funicular, a cable line. There are only two trains, each three (if I remember correctly) cars, and one train always goes on the right side of the double track area and the other on the left, each using the same track both going up and going down. How? One train has inside flanges and the other outside flanges, and the switches are thus permantely aligned with the flangeways guiding in the proper directlons. I think it is now 51 years old, still operating with the original French equipment.
I was looking at my copy of Snell’s book “Early Railways” (it is a British book from 1964) and Snell shows fairly recent photos of a coal railway that had flanges on BOTH sides of the wheel (looking a bit like an automobile wheel without the tire). The advantage was that the wheels could be loose on the axles and thus gauge was not critical (at slow speeds the theoretical advantage of “loose” wheels going around curves would probably not mean much). Snell also showed a photo of the yard tracks at one of these coal plants. The complexity of the points and frogs when flanges were on both sides of the rail was mind blowing and somehow, typically British!
As of the writing of the book – 1964 – Snell wrote that some British coal roads continued to have double flanged wheels. So at least in this unusual, limited, and specialized form outside flanges lived on longer than Mark Hemphill’s posting above would suggest .
Just for the record, in very very early days they sometimes put the flange on the rail and kept the wheels flangeless – the supposed advantage being that the same vehicle or car would be used on ordinary roads. But the cast iron flanges tended to break and the idea was largely abandoned. This too is shown in Snell’s book.
Dave Nelson
For davekeppler, one arrangement I’ve seen on funiculars (Horseshoe Curve, I think) has double flanged wheels on one side and rollers on the other. The cars have the flanges wheels on opposite sides and they carry the cars past each other at the central passing siding.
Almost all funiculars (Lookout Mountain, Mount Vesuvius, etc) have a similar arrangement to that described above. The double flanged wheel has to sit on the outside rail to steer the car at the meeting point. Funiculars are usually three-rail arrangements except at the meeting point where the center rail splits so the cars can pass.
The Horseshoe Curve funicular is two rail except for the passing section.
Pardon my ignorance, but what is a “troll” (other than the green kind that wants to collect money when you cross a bridge (unless you can answer three questions)?
Zardox, “troll” is actually a double entendre. The original sense was as a verb, meaning the same as in fishing – dragging “flame bait” to try to entice people into responding irately, or stir up discussion or knee-jerk response (or whatever the ‘troll’ wanted to get people to do).
Isn’t it like the French always to be different? Thus the flange arrangement on the Carmelit. By the way, I think the base station, not far from Israel Railway;s old Mandate-era “Merkaz” or Central Station, is (or was) named “Paris”! For prober platforming at the high platforms, only two rails are used above and below the siding/