Semper VaporoGeometry is what the acorn said when it discovered it had grown up.
Be careful. With that kind of logic people may mistake you for me.
tree68 Deggesty Actually, a tangent is a trigonometric function. Sometimes it's just something we go off on...
Deggesty Actually, a tangent is a trigonometric function.
Actually, a tangent is a trigonometric function.
Johnny
Well I'm happy my Lionel trains aren't too heavy. Their fast angle wheels make them a lot easier to pull around curves.
Modeling the "Fargo Area Rapid Transit" in O scale 3 rail.
Larry Resident Microferroequinologist (at least at my house) Everyone goes home; Safety begins with you My Opinion. Standard Disclaimers Apply. No Expiration Date Come ride the rails with me! There's one thing about humility - the moment you think you've got it, you've lost it...
Geometry is what the acorn said when it discovered it had grown up.
Semper Vaporo
Pkgs.
I thought trigonometry is geometry.
ChuckCobleighAs opposed to tangent grease
And I thought a tangent was a geometric function.
mudchicken Curve grease, properly applied (and the heck with the whining coming from operating's pie-hole)
Curve grease, properly applied (and the heck with the whining coming from operating's pie-hole)
As opposed to tangent grease as applied in "Emperor of the North" for instance?
Easy for you to say...
I've been stuck on a hill with "natural" grease (wet fall leaves) on the tracks...
selectorSurely mudchicken or PWN could clean this up for me, including correcting it largely.
Actually, I find your explanation pretty clear and I appreciate it. I'm a layman when it comes to this stuff and I never thought of the issues that arise just from a curve in the tracks.
There is a limit to how much curve a wheel set can maintain equilibrium in, but as the curve changes or as the gauge varies, the rotational speed of the wheel/axle set will change as the set moves in or out along the radius of the curve to keep the distance traveled the same as the rest of the train.
If the gauge widens, the wheel set will just spin faster to make up for the smaller diameter of both tapers and if the gauge narrows, the wheel set will spin slower due to the larger diameters.
As the curve tightens, the distance to travel on the outer rail is longer and so the wheel on that side must either spin faster or have a larger diameter to travel that longer distance without sliding... at the same time, one might consider that the inner rail is shorter and so that wheel would need to spin slower, or have a smaller diameter to traverse the distance without "peeling out".
Since the two wheels do not spin at different speeds, the Wheel Set will move out along the radius to make the outer wheel have a larger diameter at the flange side of the taper to the tread and the inner wheel have a smaller diameter and wheel set will spin at the rate that keeps the distance traveled by each wheel different by the effective lengths of the two rails.
As the curve tightens a point will be reached where the taper cannot produce the difference in diameters to keep the wheels from either sliding or overspinning. That is when the fillet of the tread to flange comes into play because the diameter rapidly increases in that small lateral distance to provide a greater distance traveled per rotation. There is also a much sharper taper to the outer edge of the tread, but at that point there is little margin for error to keep the tread on the rail and not fall in the gauge.
Somewhere I have the math worked out for what radius curve the normal taper of the tread can handle and it is quite surprising how much curve it can take before the flange fillet comes into play. I am still looking for those files (drawings and text) and if I ever remember where they are, I'll make them available... I'd hate to have to go through that again!!!!! And not too sure I even could!
BucyrusI agree that wheel taper does not center the cars on the rails. It simply compensates for the differing running distance for the two rails when negotiating a curve.
I think that pretty much everyone has said that the taper does center the wheelset on the rails. And that goes back to discussions well before this thread. Anomolies in the wheel taper and/or rail can lead to hunting, amongst other things.
If a car is standing, no, it won't slide to center up. I'll agree with that.
But when rolling, other factors notwithstanding, it is the taper that keeps the wheels (and hence the car) centered on the rails. Go back to the demo with the paper cups for a practical exercise that proves the point.
That the taper helps compensate for differing distances travelled in curves holds water. In fact, I'll agree that the effect is limited to curves of a certain radius.
Semper Vaporo Here we go again... The "differential effect" of tapered wheels would not occur UNLESS the wheels are sync'd on the same axle. If they "free wheel" then the taper is totally and completely useless. The car will not "slide sideways" to center itself between the rails because of the taper. The cars are way too heavy and the taper way too shallow for gravity to pull the metal wheels sideways on the metal rail. Also, in a curve, they will reach an equilibrium point even if the gauge varies. They will adjust the contact points on both sides until both wheels are traveling the same distance for the given diameter of the wheels.
Here we go again...
The "differential effect" of tapered wheels would not occur UNLESS the wheels are sync'd on the same axle. If they "free wheel" then the taper is totally and completely useless.
The car will not "slide sideways" to center itself between the rails because of the taper. The cars are way too heavy and the taper way too shallow for gravity to pull the metal wheels sideways on the metal rail.
Also, in a curve, they will reach an equilibrium point even if the gauge varies. They will adjust the contact points on both sides until both wheels are traveling the same distance for the given diameter of the wheels.
My point was that the freewheeling would create the differential effect as an alternative to the tapered wheels. I was not suggesting that the free wheels be combined with wheel taper.
I agree that wheel taper does not center the cars on the rails. It simply compensates for the differing running distance for the two rails when negotiating a curve.
I disagree that both wheels will find equilibrium in running distance when negotiating a curve. An equilibrium effect can only be achieved in a curve where the wheel taper/diameters are in synch with a specific curve radius, while the outside wheel flange fillets are contacting the inside of the outer railhead. Thus the wheel taper only approximates a differential effect.
Other sources say that the taper keeps the car centered on the rails and that the flanges are only for backup. I disagree with that. It is fair to say that most of the flange does not engage the rail unless the flange is badly worn. Normally, it is only the fillet radius at the base of the flange that contacts the rail. A worn flange that makes a face contact against the side of the railhead creates lift that can cause the flange to climb over the rail.
Each axle on a car might be rotating at a different speed, but on one axle both wheels must rotate together.
If I understand it, the general point of this thread was to ask if the differential action of the tapered wheels would be better accomplished by a fixed axle with independently rotating wheels. Part of the reason for the question is that the fixed, tapered wheels occasionally squeal on curves.
One big advantage I see for the fixed, tapered wheels is that an axle only requires two bearings, whereas independently rotating wheels each require two bearings, so four bearings per axle. The outboard bearings on the existing system are easy to get at and change compared to inner and outer bearings needed for a wheel rotating on a fixed axle.
In any case, the differential effect of the tapered wheels cannot be perfect because perfect geometry could only be obtained for a given curve radius. So the variation in curve radius can only match on average with the changing diameters of the wheel tapers. So, to the extent that the match is imperfect, some slippage and friction will occur between the wheels and rails.
Furthermore, if a curve is pulled too slowly, the cars will stringline to the inside rail rather than throw out to the outside rail, thus reversing the differential effect and exaggerating the original wheel speed problem for curves.
And then as Selector described earlier, you have the problem that the centerlines of the two axles of each truck in a curve do not intersect the centerpoint of the curve radius. This too causes slippage between the wheels and rails.
So the tapered wheel principle is not perfect, but good enough without resorting to more complicated and costly wheel speed differentiating systems that might be applied.
tree68 I recall seeing somewhere a while back a RR that had a wheelset change down to a science, even mid-train as Ed mentions. A couple of jacks, a forklift and a little time and they were done.
http://www.youtube.com/watch?v=MI-RjQdXToY
"At massive Bailey Yard, every minute counts for coal trains. And UP is saving hundreds of hours with in-train wheel replacements" - http://www.progressiverailroading.com/blogs/default.asp?BlogID=472&Blog=Rails%20Blog
http://en.wikipedia.org/wiki/Bailey_Yard
http://www.progressiverailroading.com/class_is/article/Dwelling-on-the-positive-at-UPs-Bailey-Yard--16467#
- Paul North.
No need for a differential.
"I have no idea why you brought up spider gears and a differential,,,, no need for them in any way shape or form."
I believe he was just making a comparison between the simplicity of one and the complexity of the other.
Norm
One interesting thing I noticed about train wheels is that the flanges on both ends of the wheels tend to stay about a inch or so away from the rails when in motion at slower speeds. The slight incline in the train wheels keeps the wheels centered using gravity. This is similar to a roulette ball spinning in roulette wheel. It takes quite a bit of speed to keep that ball up high. When it slows gravity pulls it towards it's lowest point.
With a train wheel since the wheels have a slight angle to them, which angles towards the center, the wheels will always be finding the path of least resistance. The path of least resistance seems to be right around the middle of the wheel with the flanges around 1" from the side of the rail. It takes a lot of force to push the cars wheels up that slight incline and hold them there because they naturally want to drop down due to gravity. You can actually look at brand new rails and check them after a train goes by and there will only be around 1/4" of a inch wear pattern on the head of the rail. The rest will be rusty. As time goes by the entire head will be worn.
You could probably eliminate flange wear altogether by making a train wheel twice as wide and putting a much steep angle on the wheels. If the flanges were 6 inches from the side of the rail it would take a huge amount of force to get the creep until the flanges hit the rail even at high speeds.
If you want to see some serious rail wear, the rails on the Moffat route take a serious beating. New rails around the curves are only good for around 4 or 5 months before half of the rail is ground into powder. I have seen them using the same rail but turning the bad side to the outside. In a few months that rail looks like the tip of a arrow. That environmentally friendly grease is totally worthless.
What I was imagining was each car wheel to have its own large axle shaft. Wheel on side A has the female shaft extenting say 80% of the distance towards the other wheel. Wheel on side B has the male shaft that also extends about 80% of the distance towards the other wheel. Some kind of holder could be built to keep the gauge of the axle correct. I think this could be done with existing trucks and bearings that mount on the outer ends of the wheels. OR, how about a design like the full floater axle on the rear of my 94 F250 heavy duty? I'm not a mechanical engineer but I have done quite a bit of car and truck mechanical work. If anyone would know if solid axles make a lot of drag in curves it would be the railroads. I would not be surprised if one of the RR equipment companies have experimented with trucks that have 4 wheels that can turn independantly from each other.
Try a few bits here:
Squealing while a curve is traversed has at least two generaters: one the flange scraping the outer or inner rail, and the inability of one wheel to go faster, or slower, than the other because each is traveling a distance different than the centerline of the track curve making the wheels scrape off their surfaces.
Cuesta, SP's Coast Div grade between SLO and Santa Margarita, had about a 6 month interval between new rail-transpose because of flange-wear, renew. 1 year! In the early 1960's.
Wheel slip and wheel spin....slip is bad, no traction, fall on your caboose. spin tho' is caused by at least, some, more torque than is sustainable against available wheel adhesion to the rail. GP60's ran right up to and slightly beyond that limit which generated squealing that caused me to close the cab window during a great early autumn sunset on Cuesta.
Hunting. defined as a freight car truck's wheels bouncing from one rail to the oppposite rail doesn't validate. Huntings phenomana originate at the end of car masses of bulkhead flat cars. When one of the bulkheads drops into a low rail joint it twists, torques, the other end, which not connected to the other end absorbs and then rebounds the energy....whipping each end of the car...twist a snake rapidly.
Centerbeam cars connected at the top line up uniformly with cars connected at the high portions...hunt less.
there's lots more track-train-dynaamics discovered or proven during the early 1970's.
Radial steer trucks are beginning to appear on freight cars in Europe as a method of reducing noise. Typical US type three piece freight trucks are not popular in Europe, or at least not as popular because of the separation of ownership between the Freight Companies and the owners of the track. The owners of the track tend to charge the owners of the freight cars charge more per mile operating fees versus what is charged for freight cars equipped with radial steer trucks which cause less rail wear.
I recall an ad from the 1980 Car & Loco Cyclopedia for radial freight trucks, i.e. where the axles are steered when the truck swivels. The claims were for lower rolling resistance and less hunting at speed, but my impression is that the the benefits weren't worth the extra cost. Maintenance was likely to be a problem as well.
- Erik
John WR selectorSo, if we allow that one of the axles is parallel with the radius of the curve at any one point, but also aligned with the line defining the radius at that point, the next axle(s) cannot be aligned with that radius, nor with the radius at their point(s) along the curve, they being separated from the first axle by the distance between the centres of those axles. I appreciate your clear explanation of why two axles on a truck cannot possibly be correctly aligned with the radius of a curve. However, it seems to me rather than having one axle aligned and the other not aligned both axles would be somewhat out of alignment and there would be crabbing by all four wheels. Is this true?
selectorSo, if we allow that one of the axles is parallel with the radius of the curve at any one point, but also aligned with the line defining the radius at that point, the next axle(s) cannot be aligned with that radius, nor with the radius at their point(s) along the curve, they being separated from the first axle by the distance between the centres of those axles.
I appreciate your clear explanation of why two axles on a truck cannot possibly be correctly aligned with the radius of a curve. However, it seems to me rather than having one axle aligned and the other not aligned both axles would be somewhat out of alignment and there would be crabbing by all four wheels. Is this true?
I agree, and only used my supposition to make it clear that two separated but parallel axles can't possibly be both running tangentially, and that therefore at least one of them at any point in time while traversing a curve must be crabbing. I don't see why the difference can't be split, as you suggest, and that the resulting friction can't also be split. The truth is likely to be that the trucks hunt a bit, or that their wheelsets do, so that there is a cyclic or harmonic oscillation (for want of a more educated (me) term, as I am not an engineer). Surely mudchicken or PWN could clean this up for me, including correcting it largely.
Crandell
I recall seeing somewhere a while back a RR that had a wheelset change down to a science, even mid-train as Ed mentions. A couple of jacks, a forklift and a little time and they were done.
Boyd,
The solid axles/wheelset is an integral part of the structure of the truck, they hold the side frames parallel and ridged.
It is also an easy component to replace.
If you had each wheel free turning on its own spindle and bearing, the trucks would have to be massive, with the truck and side frame carrying all the lateral loading and each wheel would need to be independently sprung, and the braking system you need for that would be complicated.
As it stands, the brakes, when applied, share some of the braking effort through the axle to the opposite side…if the left front brake shoe is a little more worn than the right front, the equalizer beam will still apply the worn brake, and some of the forces from the “good” shoe will help stop the entire axle.
Even with the entire left front shoe worn to the backing pad, the good right shoe will still stop both sides or both wheels because they are connected by the axle.
The current wheel bearing size is 6 ½ by 12 tapered roller bearing, held in place by a bearing cap, it is sealed and permanently lubed.
Remember, the simpler the design, the easier it is to repair or replace, and more often than not, the part is sturdier and longer lived.
As it is now, a good car department can swap a wheel set on just about any car in 20 minutes to a half hour, loaded or empty.
In the case of a “hot” unit train, they can do it with the car still “en train” and not even bust the air.
Say you have a bad bearing on L4 wheel…if each wheel was on its own spindle bearing system, you would have to jack up the car, remove the bearing, find a way to press the old bearing out, then press the new bearing into the wheel, the re-install the wheel on the car.
If you look at how they change a wheelset, you would find that all they have to do is jack up the car, the wheelset, bearing and bearing keepers stay on the track as you lift the car/truck off of it…roll the wheelset away, roll a new set under the car, set it down and away you go.
If each wheel was on its own spindle/bearing set, you would also have to have a way to adjust the caster and chamber on each wheel….an alignment so to say, just like on you automobile, you would have to adjust the way each wheel meets the rail head, and the angle each wheel “attacks” the track…and when you loaded the car, that would change.
If US roads went to the lighter smaller cars like most European roads have, you could make a solid truck with fixed spindles at each corner and “spring” the entire truck…chamber and caster would not be that much of a problem with the lighter car loadings, although with a solid truck and fixed spindles, crabbing in curves would still occur.
23 17 46 11
Our community is FREE to join. To participate you must either login or register for an account.